2016 Better Spending, Better Services A Review of Public Finances in Haiti 2016 BETTER SPENDING, BETTER SERVICES Better Spending, Better Services A Review of Public Finances in Haiti Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750- 8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Photo cover Credit: Isabelle Schaefer / World Bank BETTER SPENDING, BETTER SERVICES Acronyms and Abbreviations AFD Agence Française de Développement (French agency for development) AIDS Acquired Immunodeficiency Syndrome ARI Acute Respiratory Infection ASC Agent de Santé Communautaire (community health worker) BM Banque Mondiale BMPAD Bureau de Monétisation des Programmes d’Aide au Développement (Monetization Office of Development Assistance Programs) BRH Banque de la République d’Haïti (Haiti Central Bank) BSEIPH Bureau du Secrétaire d’Etat à l’Intégration des Personnes Handicapées (Office of the Secretary of State for Integration of Persons with Disabilities) CAL Centre de Santé avec Lit (health centre with bed) CAS Caisse d’Assistance Sociale (Social Assistance Fund) CDB Caribbean Development Bank CHE Catastrophic Health Expenditure CIDA Canadian International Development Agency CNMP Commission Nationale des Marchés Publics (National Procurement Commission) CONATEL Conseil National de Télécommunication (National Telecommunication Regulator) CSAFP Conseil Supérieur de l’Administration et de la Fonction Publique CSCCA Cour Supérieure des Comptes et du Contentieux Administratif (Supreme Court of Accounts and Contentious Administrative Proceedings) CSL Centre de Santé sans Lit (health centre without bed) DAB Directorate of Administration and Budget DAO Dossier d’Appel d’Offres (Bidding Document) DASIP Direction de l’Analyse et du Suivi des Investissements Publics (Public Investment Analysis and Monitoring Unit) iii DDP Document Définitif du Projet (Final Project Document) DEA Data Envelopment Analysis DEC Direction d’Evaluation et de Contrôle (Evaluation and Control Directorate) DGB Direction Générale du Budget (Budget Directorate) DTDCP Direction du Trésor, Dette, et Comptabilité Publique (Treasury Directorate) DHS Demographic and Health Survey DIP Direction de l’Investissement Public (Directorate of Public Investment) DPC Direction de la Pension Civile (Directorate for Public Pensions) DPES Direction de la Programmation Economique et Sociale (Economic and Social Programming Directorate) DSE Direction du Suivi et de l’Evaluation (Directorate of Monitoring & Evaluation) ECVMAS Enquête sur les Conditions de Vie des Ménages après le Séisme (survey on the living conditions of households after the earthquake – household survey) EMMUS Haiti Mortality, Morbidity, and Service Utilization Survey (DHS in English) EPPLS Entreprise Publique de Production de Logement Social (Public Enterprise for the Promotion of Social Housing) EPSSS Evaluation de la Prestation des Services de Soins de Santé (Service Provision Assessment) EPT Education Pour Tous (Education For All) EU European Union FAES Fonds d’Assistance Economique et Social (Economic and Social Assistance Fund) FER Fonds d’Entretien Routier (Road Maintenance Fund) FIOP Fiche d’Identification et d’Opération des Projets (Identification and Project Operation File) FNE Fonds National de l’Education (National Education Fund) FY Fiscal Year GDP Gross Domestic Product GL General Ledger GoH Government of Haiti GPRSP Growth and Poverty Reduction Strategy Paper GSP Groupe Santé Plus (Health group – private company) HCR Hôpital Communautaire de Référence (Community Referral Hospital) HIMO Travaux à Haute Intensité de Main d’Oeuvre (Labor-intensive public works) HIS Health Information System HIV Human Immunodeficiency Virus HRMIS Human Resources Management Information System HTG Haitian Gourdes IADB Inter-American Development Bank IBERS Institut du Bien-Etre Social et de Recherches (Institute of Social Welfare and Research) IBRD International Bank for Reconstruction and Development iv BETTER SPENDING, BETTER SERVICES IDA International Development Association IDB Inter-American Development Bank IGF Inspection Générale des Finances (Inspectorate General of Finances) IHE Institut Haïtien de l’Enfance (Haitian Institute of Childhood) IHSI Institut Haïtien de Statistique et d’Informatique (Haitian Statistics Agency) IMF International Monetary Fund IMR Infant Mortality Rate INTOSAI International Organization of Supreme Audit Institutions KF Kore Fanmi (Family Support – Community Social Worker Initiative) LdF Loi de Finances MARNDR Ministère de l’Agriculture, des Ressources Naturelles et du Développement Rural (Ministry of Agriculture, Natural Resources and Rural Development) MAST Ministère des Affaires Sociales et du Travail (Ministry of Social Affairs and Labor) MCFDF Ministère de la Condition Féminine et des Droits des Femmes (Ministry for Women and Women’s Rights) MDG Millennium Development Goals MEF Ministère de l’Economie et des Finances (Ministry of Economy and Finance) MENFP Ministère de l’Education Nationale et de la Formation Professionnelle (Ministry of National Education and Vocational Training) MICT Ministère de l’Intérieure et des Collectivités Territoriales (Ministry of Interior and Local Government) MIS Management Information Systems MMR Maternal Mortality Ratio MNCH Maternal, Neonatal, and Child Health MoP Ministry of Planning MPCE Ministère du Plan et de la Coopération Extérieure (Ministry of Planning and External Cooperation) MSPP Ministère de la Santé Publique et de la Population (Ministry of Public Health and Population) MTPTC Ministère des Travaux Publics, Transports et Communications (Ministry of Public Works, Transport, and Communication) NGO Non Governmental Organization NHA National Health Account OFATMA Office d’Assurance Accidents du Travail, Maladie et Maternité (Office of Insurance for Work Accidents, Illness and Maternity) OMRH Office des Ressources Humaines (Office of Human Ressources) ONA Office National d’Assurance Vieillesse (National Office for Old Age Insurance ) ONART Office National de l’Artisanat (National Handicrafts Office) ONM Office National de la Migration (National Migration Office) PAARP Plan d’Action pour l’Accélération de la Réduction de Pauvreté (Plan for Accelerating the Reduction of Extreme Poverty) v PaP Port-au-Prince PARDH Plan d’Action pour le Relèvement et le Développement d’Haïti (Action Plan for Haitian National Recovery and Development) PDS Plan Directeur de Santé (Health Master Plan) PEFA Public Expenditure and Financial Accountability PFM Public Financial Management PIMS Public Investment Management System PIP Programme d’Investissement Public (Public Investment Program) PMS Paquet Minimum de Santé (Minimum Package of Services) PNCS Programme National de Cantines Scolaires (National School Canteen Program) PNS Politique Nationale de Santé (National Health Policy) PSDH Plan Stratégique de Développement d’Haïti (Strategic Plan for the Development of Haiti) PSUGO Programme de Scolarisation Universelle Gratuite et Obligatoire (Free and Compulsory Universal Schooling Program) PTI Plan Triennal d’Investissement (Triennial Investment Plan) RBF Results-Based Financing SAI Supreme Audit Institution SDI Schéma Directeur Informatique (IT Master Plan) SIGMP Système Informatisé de Gestion des Marchés Publics (IT System for Procurement Management) SP Social Protection SSN Social Safety Net SYSCOMPTE Système de gestion des Comptes Courants (Accounting System for Specific Budget Items) SYSDEP Système Informatisé de gestion des Dépenses Publiques (Public Expenditure Management System) SYSPIP Système Informatisé de gestion du Programme d’Investissement Public (Public Investment Management System) TB Tuberculosis THE Total Health Expenditure U5MR Under-five Mortality Rate UAS Unité d’Arrondissement de Santé (District Health Unit) UCE Unité de Contrôle d’Exécution UEP Unité d’Etude et de Programmation (Programming and Analysis Unit) ULCC Unité de Lutte Contre la Corruption (Anti-Corruption Unit) UN United Nations UNCITRAL United Nations Commission on International Trade Law UNDP United Nations Development Program UNICEF United Nations Children’s Fund UPM Unité de Passation de Marché (Procurement Unit) vi BETTER SPENDING, BETTER SERVICES USAID United States Agency for International Development WB World Bank WFP World Food Program WHO World Health Organization WHS World Health Survey vii viii BETTER SPENDING, BETTER SERVICES Acknowledgement W e would like to thank the members of the Haiti Country Team, as well as all the partners and stakeholders in Haiti who have contributed to the the following institutions: the Ministry of Economy and Finance, the Ministry of Education, the Ministry of Planning and External Cooperation, the Ministry preparation of this document in a strong collaborative of Public Health and Population, the Ministry of process. We are very grateful for the generosity Social Affairs, and the Prime Minister’s Office, as exhibited in providing us with substantive inputs, well as the Central Bank of Haiti (BRH). The Team knowledge and advice. The Team was led by Raju Jan wishes to thank all the participants of the numerous Singh (Program Leader, LCC8C) and the table below workshops where preliminary results have been identifies the full list of team members who have presented and discussed, in particular Mme Laleau, contributed their time, effort and expertise, and their Head of the Economic Studies (Ministry of Economy affiliations. and Finance) and Berny Duvalsaint, Secretary of the PER Steering Committee. The Team gratefully We wish to thank for their helpful suggestions acknowledges also the support from donors present and insights our peer reviewers: Margaret Grosh in Haiti: AFD, Brazil, Canada, European Union, (Practice Manager, GSPDR), Sebastian James IADB, IMF, Mexico, USAID, UNDP, and UNICEF. (Senior Investment Policy Officer, GTCDR), Roland Kpodar (Senior Economist, IMF), David Cal A strong emphasis has been put throughout the McWilliam (Senior Economist, GMFDR), Patrick preparation of this report on training and capacity Ramanantoanina (Senior Operations Officer, building. Special attention has been put on ensuring GEDDR), and Anand Rajaram (Practice Leader, as much as possible that the various analytical tools GGODR). Mary Barton-Dock (Special Envoy for and technics used in this document be transferred Haiti, LCC8C) and Miria Pigato (Practice Manager, to our Haitian counterparts, and many of the GMFDR) advised the Team throughout the stages of quantitative work was carried out jointly with Haitian the process. teams. In this regard, the Team would like to thank particularly all the Haitian participants who attended The Team is also extremely grateful for the our various workshops on the Tariff Reform Impact cooperation of the Haitian authorities and for the Simulation Tool (TRIST), the subsidy simulation invaluable contributions of senior officials and toolkit (SUBSIM), and BOOST, as well as Ibrahim El governmental agencies. We acknowledge in particular ghandour (GGODR), Olivier Jammes (Consultant, ix GEDDR), and Paulo Verme (Senior Economist, Data in Haiti is particularly challenging. This study GPVDR) for running these trainings. built on the recently-completed Poverty Assessment, carried out jointly by the World Bank and the The Team is also grateful to a number of colleagues ONPES. The expansion in donor assistance and the who provided insights, comments and support in availability of concessional financing over the past the course of the report preparation. They include decade has compounded, however, the fragmentation Lucy Basset (Social Protection Specialist, GSPDR), of fiscal data. In this regard, the Team would like to Malaika Becoulet (Consultant, GTIDR), Ludmilla thank the Gates Foundation financing the BOOST Buteau (Consultant, GHNDR), Pierre Bonneau Initiative in Haiti, as well as Massimo Mastruzzi (Program Leader, LCC8C), Daniel Boyce (Practice (Senior Economist, GGODR) and Leif Jensen (Senior Manager, GGODR), Michelle Keane (Lead County Public Sector Specialist, GGODR) for assisting the Officer, LCC8C), Donald Mphande (Lead Financial Team in this process. Our work on fuel price subsidies Management Specialist, GCFDR), Deo Ndikumana would not have been possible without the financial (Senior Country Officer, LCC8C), Prosper Nindorera support of the Energy Sector Management Assistance (Senior Procurement Specialist, GGODR), Maki Program (ESMAP), and the help of Sameer Shukla Noda (Consultant, GSPDR), Elizabeth Ruppert (Senior Energy Specialist, GEEES) and Rohit Khanna Bulmer (Lead Economist, GCJDR), Paolo Verme (Practice Manager, GEEES). (Senior Economist, GPVDR), and Kanae Watanabe (Country Officer, LCC8C). HAITI PER TEAM Global Practice/Cross-cutting Team Members Area Communication Christelle Chapoy, Berdine Edmond Education Melissa Adelman, Juan Baron, Eva Junyen, Axelle Latortue Energy & Extractives Frederic Verdol Governance Emeline Bredy, Mamadou Deme, Ibrahim El ghandour, Eduardo Estrada, Rubens Lacerd, Andy MacDonald, Fabienne Mroczka, Renas Sidahmed, Gerard Verger Haiti CMU Gabrielle Dujour, Nellie Sew Kwan Kan, David Lighton, Raju Jan Singh, Paula White Health, Nutrition & Population Eleonora Cavagnero; Marion Cros; Sunil Rajkumar Macro Economics & Fiscal Kassia Antoine, Calvin Djiofack, Evans Jadotte, Naoko Kojo, Jan Loeprick, Management Julie Lohi, Sandra Milord, Erik von Uexkull, Konstantin Wacker Poverty Facundo Cuevas, Federica Marzo, Aude-Sophie Rodella, Thiago Scot Social Protection Carine Clert, Victoria Strokova, Frieda Vandeninden Social, Urban, Rural and Michel Matera, Rafael Van der Borght Resilience x BETTER SPENDING, BETTER SERVICES xi Emmanuel Wilkinton fait ses devoirs de sciences à Delmas 32 , Haïti. Emmanuel est l’un des bénéficiaires ayant accès à l’électricité jusqu’à 23 heures et peut maintenant étudier facilement chaque soir. Photo : Dominic Chavez / Banque Mondiale xii BETTER SPENDING, BETTER SERVICES Contents Acronyms and Abbreviations.......................................................................... iii Acknowledgement.............................................................................................. ix Executive Summary............................................................................................. xxv CHAPTER 1 Country Profile........................................................................... 1 A. Haiti Has Opportunities …......................................................................................... 1 B. … Hampered by Political Instability and Natural Disasters................................... 2 C. … As Well as Weak Structural Policies and Institutions......................................... 2 CHAPTER 2 Living With Tighter Budget Constraints.......................... 9 A. Macroeconomic Stability Has Been Maintained …................................................. 9 B. … With Greater Revenue Mobilization and a Surge in Aid.................................... 10 C. … Allowing Higher Public Spending......................................................................... 12 D. … Despite Widening Deficits, Debt is Expected to Remain Sustainable …......... 13 E. … But Financing Constraints Are Becoming Tighter............................................. 17 F. … While Natural Disasters Would Call for Larger Fiscal Buffers.......................... 17 CHAPTER 3 Protecting Priority Spending................................................. 19 A. Haiti Has a Vision Calling for Higher Human and Physical Capital …................ 19 B. … Reflected in the Rise in Public Investment …..................................................... 24 C. … And Greater Priority Given to Social Sectors …................................................. 26 xiii D. … But Haiti’s Inclusiveness Remains Limited........................................................... 29 CHAPTER 4 TOWARDS GREATER AND MORE EQUITABLE REVENUE MOBILIZATION......................................................... 31 A. Despite Some Increase, Revenue Mobilization Remains Low and Regressive….31 B. … With Exemptions and Ill-Designed Tax Brackets Eroding Direct Taxes…..... 34 C. … And Indirect Taxes Hampered by Weak Revenue Administration and Unclear Exemptions.............................................................................................. 40 D. … But Could Be Bolstered By a Simple VAT............................................................ 42 E. … And Taxes on International Trade Could Be Streamlined................................. 44 CHAPTER 5 GETTING A GREATER GROWTH DIVIDEND FROM PUBLIC INVESTMENT............................................................................ 51 A. High Public Investment Doesn’t Translate into Faster Growth in Haiti…............ 51 B. … Because of an Overly Elaborate and Rarely Respected Regulatory Framework................................................................................................. 56 C. … Little Strategic Guidance to Select Projects.......................................................... 57 D. … A Budgeting Process That is Not Followed.......................................................... 60 E. … Weak Procurement….............................................................................................. 63 F. … Fragmented Project Execution and Administration…....................................... 65 G. … Limited Physical and Financial Monitoring….................................................... 69 H. … And a Lack of Ex-post Evaluation (Project Auditing and Evaluation)............. 72 CHAPTER 6 Increasing Sustainability and Improving Further Health Outcomes............................................................................... 75 A. Although Low, Health Indicators Have Improved ….............................................. 75 B. … But Disparities Remain …...................................................................................... 79 C. … With Relatively High Health Spending Financed by Donors............................ 81 D. … Access to Health Is Limited Especially For the Poor …..................................... 83 E. … And Public Spending Is Not Geared to Better Services...................................... 85 CHAPTER 7 KEEPING CHILDREN IN SCHOOL AND IMPROVING EDUCATION OUTCOMES............................................... 99 A. Despite Progress, Education Outcomes are Low …................................................. 99 B. … Driven by a Greater Education Supply from the Non-Public Sector ….......... 103 xiv BETTER SPENDING, BETTER SERVICES C. … While Public Spending Has Been Recently Rising …........................................ 104 D. … Its Composition Is Not Conducive to Better Services …................................... 107 E. … With Cost Remaining a Main Obstacle to Access…........................................... 109 F. … And Disparities across Regions and Poor Services............................................. 110 CHAPTER 8 TOWARDS A MORE EFFECTIVE AND BETTER TARGETED SOCIAL PROTECTION SYSTEM....................... 113 A. Growing Investments in Social Safety Nets............................................................... 113 B. … Needed By A Vulnerable Population …............................................................... 117 C. … But Too Fragmented ….......................................................................................... 120 D. … Too Small …............................................................................................................. 122 E. … Not Well Targeted …............................................................................................... 127 F. … And of Limited Impact........................................................................................... 131 CHAPTER 9 PREVENTING A RETURN OF FUEL SUBSIDIES.................... 135 A. Fuel Price Subsidies Implied a Heavy Burden on the Budget…............................. 135 B. … Are Regressive…...................................................................................................... 138 C. … And Could Return................................................................................................... 146 Concluding Observations................................................................................. 149 References............................................................................................................. 153 xv List of Annexes Annex 1 : PIM Process Flowcharts................................................................................. 166 Annex 2 : Correlates with Enrollment............................................................................ 171 Annex 3 : Correlates with Overage Status...................................................................... 172 Annex 4 : EDE PEP Programs: A Glossary.................................................................... 173 Annex 5 : Price Structure for Petroleum Products (2010/2011)................................. 175 List of Boxes Box 2.1 : Electricité d’Haiti............................................................................................... 15 Box 2.2 : Petrocaribe and Haiti........................................................................................ 16 Box 3.1 : Benchmarking the Drivers of Shared Prosperity: An Application to Haiti.................................................................................... 22 Box 3.2 : Fiscal Data Challenges...................................................................................... 27 Box 4.1 : International Experience with Tax Incentives............................................... 35 Box 4.2 : Removing Nuisance Taxes................................................................................ 37 Box 4.3 : The CATT Methodology for Benchmarking Customs Performance......... 42 Box 4.4 : The World Bank’s Tariff Reform Impact Simulation Tool (TRIST)............ 48 Box 5.1 : Common Features of PIM in Donor-Dependent Countries....................... 54 Box 5.2 : Haiti Legislative Framework............................................................................ 56 Box 5.3 : The Share of Recurrent Expenditures in Investment Projects..................... 61 Box 5.4 : INTOSAI Conditions Supporting Corruption.............................................. 65 Box 5.5 : Project Completion........................................................................................... 74 Box 6.1 : Health Care Provision in Haiti........................................................................ 86 Box 6.2 : Definitions of Operating Budget Terms......................................................... 88 Box 6.3 : Definition of Technical Efficiency................................................................... 94 Box 6.4 : More Health for Every Dollar: Results-Based Financing............................. 96 Box 6.5 : The Right Incentives Lead to Measurable Results in Rwanda..................... 97 Box 7.1 : Education Provision in Haiti........................................................................... 105 Box 8.1 : Methodology Note on Social Protection Spending...................................... 125 Box 9.1 : The Transport Sector in Haiti.......................................................................... 143 xvi BETTER SPENDING, BETTER SERVICES List of Figures Figure 1.1 : Annual GDP Growth vs. Change in Government, 1971-2013 (percent).................................................................................. 3 Figure 1.2 : Annual GDP Growth vs. Occurrence of Natural Disaster, 1971-2013 (percent).................................................................................. 3 Figure 1.3 : Obstacles to Growth................................................................................... 4 Figure 1.4 : Change in Governance Indicators, 2004-13........................................... 6 Figure 1.5 : Logistic Performance Index, 2014............................................................ 7 Figure 1.6 : Port Tariffs Estimated Cost Per TEU*, 2009 (in US Dollar)................. 7 Figure 1.7 : Electric Power Consumption, 2011......................................................... 8 Figure 2.1 : Inflation, 2004-14....................................................................................... 10 Figure 2.2 : Gross International Reserves, 2004-14 (Millions of US Dollars)......... 10 Figure 2.3 : Fiscal Revenues, 2004-13 (In Percentage of GDP)................................ 12 Figure 2.4 : General Government – Total Revenues, 2000-12 (In Percentage of GDP)............................................................................. 12 Figure 2.5 : Public Expenditures by Source of Financing, 2001-13 (In Percentage of GDP)............................................................................. 13 Figure 2.6 : General Government Expenditure - International Comparison 2004-13................................................................................. 13 Figure 2.7 : Central Government Fiscal Balance,....................................................... 14 Figure 2.8 : Current Account Balance, 2004-14 (Percentage of GDP)..................... 14 Figure 2.9 : Haiti – Stock of Debt from Petrocaribe, 2009-14 (In Percentage of GDP)............................................................................. 14 Figure 2.10: External Debt (PV), 2014-35 (Percentage of Exports G&S + Remittances)......................................... 14 Figure 2.11: International Aid, 2008-25 (In Percentage of GDP).............................. 17 Figure 2.12: Government Deposits, 2009-14 (In Percentage of GDP)..................... 17 Figure 2.13: Petrocaribe Financing, 2008-17 (Millions of US Dollars).................... 18 Figure 3.1 : Extreme Poverty Simulations.................................................................... 20 Figure 3.2 : Economic Magnitude of Estimated Parameters..................................... 21 Figure 3.3 : Life Expectancy at Birth, 2010 (Years)..................................................... 23 Figure 3.4 : Cabinet Changes, 2003 (10 Years Average)............................................. 23 Figure 3.5 : Income Effects of Closing the Gap (In Percentage of the Gap)............ 23 xvii Figure 3.6 : Current Expenditure, 2005-14 (In Percentage of GDP)........................ 26 Figure 3.7 : Operating and Capital Expenses, 2005-14 (In Percentage of GDP).... 26 Figure 3.8 : Changes in Sectoral Composition –........................................................ 28 Figure 3.9 : Project Activities Financed by Petrocaribe Funds, 2008-13, (USD Millions)........................................................................................... 28 Figure 3.10: Project Activities Financed by Petrocaribe Fund, 2008-13, (Percentage of Total)................................................................................. 29 Figure 3.11: Donors Financing, 2010-12 (Percentage of Total)................................. 29 Figure 3.12: International Comparison - Social Spending, 2013 or latest (In Percentage of GDP)............................................................................. 30 Figure 4.1 : Fiscal Revenue, 2009-13 (In Percent of GDP)........................................ 32 Figure 4.2 : Tax-to-GDP Ratio...................................................................................... 33 Figure 4.3 : Ratio Direct To Indirect Taxation, 2009 or 2011.................................... 34 Figure 4.4 : Corporate (left) and Personal (right) Income Tax Rates in Regional Comparison (Percent).......................................................... 35 Figure 4.5 : Income Tax Rate Thresholds In Regional Comparison (Per Capita GDP)....................................................................................... 38 Figure 4.6 : Hypothetical Income Tax Revenue Collection Based on 2012 Household Survey Results........................................................................ 39 Figure 4.7 : Hypothetical Effect of Changes in Income Tax Brackets on Revenue Collection.............................................................................. 40 Figure 4.8 : Results of Haiti Customs Assessment: Distance from........................... 41 Figure 4.9 : Tax Revenue Collection around the Introduction of VAT.................... 43 Figure 4.10: Applied Protection from Tariffs and Inspection Fees for Four Reform Scenarios....................................................................... 48 Figure 5.1 : Capital Investment (a) and (b).................................................................. 52 Figure 5.2 : Public Management Efficiency Index...................................................... 53 Figure 5.3 : Public Investment Management Index – PIM (0=Lowest, 4=Highest)............................................................................. 53 Figure 5.4 : Four Stages of the Project Life Cycle........................................................ 55 Figure 5.5 : Required Process for Adding New Projects to the PIP......................... 59 Figure 5.6 : Illustrates the Budget Formulation Process for the Preparation of the Investment Budget............................................. 62 Figure 5.7 : Contracting Methods, 2012-13................................................................. 63 xviii BETTER SPENDING, BETTER SERVICES Figure 5.8 : Complexity in Procurement Processes.................................................... 66 Figure 5.9 : PIP Project Execution Processes.............................................................. 67 Figure 5.10: Treasury funded Public Investment........................................................ 70 Figure 5.11: Cash Flows National Treasury Funded Projects 2012-13..................... 71 Figure 5.12: Monitoring and Evaluation on Projects, in PIP, 2011-13 (in Number of Projects)............................................................................ 73 Figure 6.1 : Maternal Mortality, 2013 (Per 100,000 live birth).................................. 76 Figure 6.2 : Under-5 Child Mortality, 2013 (Per 1,000 live births)........................... 76 Figure 6.3 : Health Service Utilization (In Percent of Population)........................... 77 Figure 6.4 : Changes in Under-5 Child Mortality Rates, 1990-2013 (per 1,000 live births)................................................................................ 78 Figure 6.5 : Maternal Mortality Ratio, 1990-2013 (Per 100,000 live births)........... 78 Figure 6.6 : Changes in Life Expectancy at Birth,....................................................... 78 Figure 6.7 : Health Service Utilization, 1994-2012 (In Percent of Population)..... 78 Figure 6.8 : Change in Infant Mortality, 2006-12 (Percentage Points)..................... 79 Figure 6.9 : Change in Under-5 Child Mortality, 2006-12 (Percentage Points)...... 79 Figure 6.10: Change in Stunting Rate, 2006-12 (Percentage Points)......................... 79 Figure 6.11: Change in Diarrhea Prevalence Rate, 2006-12 (Percentage Points).... 79 Figure 6.12: Child Health Outcome Indicators, 2012 (Per 1,000 Live Births)......... 81 Figure 6.13: Health Expenditure Per Capita, 2012 (Current US$)............................ 82 Figure 6.14: Health Expenditure Per Capita PPP, 1995-2012 (Constant 2005 US$)................................................................................. 82 Figure 6.15: Public Health Expenditure, 2012 (Percent of GDP).............................. 82 Figure 6.16: Public Health Spending, 2000-12 (Percent of Government Expenditure)................................................... 82 Figure 6.17: Health Expenditure, 2005-10, (Percentage of Total)............................. 82 Figure 6.18: Health Expenditure, 2011-12 (Percentage of Total).............................. 82 Figure 6.19: Health Expenditure by Source, 2004-2013............................................. 83 Figure 6.20: Barriers to Health Care Access for Women Aged 15-49, 2012 or 2013 (Percentage of Total).......................................................... 84 Figure 6.21: Enrollment of Populations in Risk-Pooling............................................ 84 Figure 6.22: Incidence of Catastrophic Health Expenditures at................................ 85 Figure 6.23: Salaries, Excl. Domestic Investment Budget........................................... 89 xix Figure 6.24: Investment Budget from Public Funds, 2012-13 (Percentage of Total)................................................................................. 89 Figure 6.25: Utilization of the Different Types of Structures by Area, 2013............ 89 Figure 6.26: Density of Medical Personnel, 2013 (Per 10,000 Inhabitants)............ 91 Figure 6.27: Density of Inpatient Beds, 2013................................................................ 91 Figure 6.28: Density of Medical Personnel by Department – Private for-Profit Sector Excluded......................................................................................... 91 Figure 6.29: Number of Beds, exc. Private For-Profit (Per 10,000 Inhabitants)...... 91 Figure 6.30: Density of (a) Medical (b) Paramedical Personnel, and (c) Community Agents by Location,............................................................ 91 Figure 6.31: Technical Efficiency Score, 45 First-Level............................................... 93 Figure 6.32: Consultations Per Doctor and Nurse, exc. Private For-Profit Sector, 2013 (Daily Numbers).................................................................. 94 Figure 6.33: Consultations per Medical Staff............................................................... 94 Figure 6.34: International Comparisons-Share of Administrative Personnel In Total Number of Primary Health Facilities (Percentage Total)...... 95 Figure 7.1 : Share of 3-25 Year Olds Currently............................................................ 100 Figure 7.2 : Haiti’s Literacy Rate Is Between LIC and LAC Averages....................... 100 Figure 7.3 : Figures (a) and (b) - Gross and Net Enrollment Ratios, 2001-12........ 101 Figure 7.4 : Youth Enrolled in School by Age and...................................................... 104 Figure 7.5 : Number of Primary Schools, 1930-2011 (Number of Schools)........... 105 Figure 7.6 : Donor Financing for Education Sector, 2010-13................................... 105 Figure 7.7 : MENFP Annual Budget Funded by Domestic Resources, 2005-13....................................................................................................... 106 Figure 7.8 : Student Beneficiaries by Department, 2013............................................ 107 Figure 7.9 : Public Expenditures on Education........................................................... 108 Figure 7.10: Education Budget, 2014 (domestic resources only, percentage of total).................................................................................... 108 Figure 7.11: Education Expenditures by Type, 2012................................................... 109 Figure 8.1 : EDE-PEP SSN Financing, By Source, 2013 (Percentage of Total)....... 116 Figure 8.2 : Programs Under EDE PEP........................................................................ 116 Figure 8.3 : Chronic Poverty, Multidimensional Deprivation and Transient Poverty, 2012 (Percent of Total)............................................................... 118 Figure 8.4 : Poverty Headcount by Age Group, 2012 (In Percent)........................... 120 xx BETTER SPENDING, BETTER SERVICES Figure 8.5 : Key Risks, the Life Cycle, and Social Protection in Haiti...................... 121 Figure 8.6 : Public Expenditure on Social Protection, 2013 (Percentage of GDP)................................................................................. 124 Figure 8.7 : Social Safety Net Spending – International Comparison, 2013 or latest (Percent of GDP)............................................................... 126 Figure 8.8 : Distribution of Social Protection Benefits (Benefit Incidence), 2012.......................................................................... 128 Figure 8.9 : Access to Social Security by Income Group, 2012 (Percentage of Total)................................................................................. 129 Figure 8.10: Coverage of Social Assistance Programs and Distribution of Beneficiaries........................................................................................... 130 Figure 8.11: Coverage of Social Assistance Programs by Age Group, 2012,............ 131 Figure 8.12: EDE PEP Coverage (FAES-Executed) by Region and Program Type, 2013........................................................................... 132 Figure 8.13: Numbers of Meals Distributed, Kantin Mobile, 2013 (Percentage of Total Population)............................................................. 132 Figure 8.14: Benefit Amounts and Contribution to Consumption of Beneficiaries........................................................................................... 132 Figure 9.1 : Evolution of Domestic Price and International Prices of Oil Products, 2004-14........................................................................... 136 Figure 9.2 : Oil Subsidies Costs, 2010-13..................................................................... 137 Figure 9.3 : Total Post-Tax Costs of Petroleum Subsidies, 2011 (Percentage of GDP)................................................................................. 137 Figure 9.4 : Total Post-Tax Costs of Petroleum Subsidies, 2011 (In percent of General Revenues)............................................................ 137 Figure 9.5 : Oil Subsidies and Public Spending Allocation, 2010-12....................... 138 Figure 9.6 : Expenditure on Petroleum Products Per Decile, 2012 (Percentage of Budget).............................................................................. 139 Figure 9.7 : Fuel Subsidies Received By Group, 2012 (Percentage of Total)........... 140 Figure 9.8 : Composition of Expenditure by Groups, by Decile, 2012.................... 140 Figure 9.9 : Expenditures by Groups, 2012.................................................................. 140 Figure 9.10: Welfare Loss by Direct Effect (Percentage)............................................. 141 Figure 9.11: Intensity In Oil Products By Sector......................................................... 142 Figure 9.12: Effect of Prices from the IO Table............................................................ 142 Figure 9.13: Contribution to Inflation........................................................................... 142 xxi Figure 9.14: Welfare Loss by Decile (Percent)............................................................. 145 Figure 9.15: Main Source of Lighting, 2012 (percent Households)......................... 145 Figure 9.16: Households Using Kerosene As Main Source of Lighting, 2012 (Percent)............................................................................................ 145 Figure 9.17: Total Loss of Welfare by Decile................................................................. 146 Figure 9.18: Welfare Losses by Urban/Rural Areas..................................................... 146 List of Tables Table 1.1 : Frequency and Impact of Natural Disasters, 1971-2014.......................... 3 Table 2.1 : Macroeconomic Indicators........................................................................... 11 Table 2.2 : Fiscal Revenues - International Comparison (In Percentage of GDP)................................................................................ 12 Table 3.1 : Fiscal Indicators (In Percent of GDP)........................................................ 25 Table 4.1 : Customs Duties and Inspection Fees (2013).............................................. 45 Table 4.2 : TRIST Simulations on Four Scenarios of Tariff Reform.......................... 47 Table 5.1 : Set of Projects Reviewed by Ministerial Affiliations................................. 55 Table 5.2 : Procurement Thresholds for State institutions (in millions of Haitian Gourdes) ................................................................ 64 Table 5.3 : Summary of Applications Relating to the Expenditures and Investment Chain.................................................................................. 71 Table 6.1 : Maternal and Children Health Coverage by Income Group, 2005-6 and 2012 (In Percent of Population).............................................. 80 Table 6.2. Participation Incidence by Income Groups, 2013 (in Percent of Beneficiaries)......................................................................... 85 Table 6.3 : Technical Efficiency In Haiti And Other Low-Income Countries........... 93 Table 7.1 : Late Starts, Repetition, and Drop out Contribute to Low System Efficiency.............................................................................. 102 Table 7.2 : National Exam Passing Rates, 2012-13 (In Percent)................................. 103 Table 7.3 : Characteristics of Primary Schools by Department, 2010-11 (In Percent)..................................................................................... 110 Table 8.1 : Social Insurance: Institutions and Programs in Haiti, December 2014.............................................................................................. 122 xxii BETTER SPENDING, BETTER SERVICES Table 8.2 : EDE PEP Fragmentation: Institutions and Programs by October 2014............................................................................................ 123 Table 8.3 : EDE PEP Executed Spending (Actual) in 2013......................................... 126 Table 8.4 : Key Donors’ Contributions (Estimates) - Social Protection.................... 127 Table 8.5 : Alignment of EDE PEP Programs with Risks and Vulnerabilities Across the Life Cycle..................................................................................... 133 Table 9.1 : Distribution of Oil Subsidies by Fuel Product, 2009-13 (Billion Gourdes)........................................................................................... 137 Table 9.2 : Average Yearly Expenditure on Fuels (HTG)............................................ 139 Table 9.3 : Fuel Subsidies Received By Income Decile (Percentage of Total)........... 140 Table 9.4 : Subsidies (Million Gourdes) Given Different Scenarios for Oil Prices.................................................................................................. 147 xxiii xxiv BETTER SPENDING, BETTER SERVICES Executive Summary Objectives GDP per capita fell by 0.7 percent per year on average between 1971 and 2013. As a result, Haiti 1. Haiti has a vision to become an emerging is the poorest country in the Latin America and economy by 2030. Haiti has comparative Caribbean (LAC) region and among the poorest advantages, including its proximity and access in the world. The overall poverty headcount to major markets; a young labor force and a amounts to about 59 percent and extreme dynamic diaspora; and substantial geographic, poverty to 24 percent in 2012, indicating that historical, and cultural assets. Areas of economic almost 6.3 million Haitians cannot meet their opportunity for Haiti include agribusiness, light basic needs and 2.5 million cannot even cover manufacturing and tourism. Recognizing these their food needs. Furthermore, with a Gini opportunities, the Government of Haiti issued in coefficient at 0.6, Haiti has the highest income May 2012 a Strategic Development Plan (PSDH), inequality in the region and one of the highest aiming at building a new modern, diversified, in the world. resilient, competitive and inclusive economy, 3. Haiti experienced a return of donor assistance respectful of its environment and in which and greater access to concessional financing people’s basic needs are met. Achieving this over the past decade. Haiti has long been objective would require ambitious double digit characterized by its very low fiscal revenue growth rates, a significant break from the past, mobilization, seriously constraining its ability possibly based on an expansion of agriculture, to carry out needed developmental spending construction, manufacturing, and tourism. (infrastructure, health, education). A lot of 2. Haiti’s growth performance in the last four basic services are provided in Haiti by non-State decades has been overall disappointing, actors. Following greater political stability and however, and poverty remains endemic. A particularly the 2010 earthquake, the budget has history of vested interests, political instability, benefitted from exceptional donor assistance and natural disasters, as well as poor governance, with external grants increasing from 2 percent inadequate infrastructure and limited skills, of GDP in 2004 to 8.1 percent in 2013, peaking have prevented the country to realize up to now at 12.1 percent of GDP in 2010. Meanwhile, in its aspirations, trapping it in a low equilibrium. addition to benefiting from the HIPC and the xxv MDRI initiatives, Haiti received additional debt specifically (e.g. equipment, utilities) have not cancellation in the aftermath of the earthquake, kept up with rising public investment, remaining reducing its total external debt to 8.9 percent of broadly unchanged over the past decade. Without GDP in 2011 and providing borrowing space the needed resources to operate the newly built that the country used for concessional financing investments and equipment, it is questionable from Venezuela. whether the new facilities could be maintained or deliver the expected growth dividends. 4. These resources allowed Haiti to finance a strong expansion in capital spending. Greater donor 7. Furthermore, despite the recent increase assistance and the availability of concessional in social public spending, Haiti’s economic financing have allowed an expansion in capital inclusiveness is still limited. Social spending spending, a substantial shift in the country’s remains limited and the delivery of basic services priorities as reflected by the growing share of highly inequitable. Public spending in health, public investment: it represented in 2014 more education, and social protection amounts to 5 than half of total public spending compared to percent of GDP, below comparator countries, a third in 2005. In line with higher own fiscal limiting the government’s ability to offer revenue, current expenditure has also been on equal opportunities to its citizens. At the same the rise, increasing from just below 10 percent of time, many large spending items such as fuel GDP in 2005 to about 13 percent GDP in 2014, subsidies clearly favor the rich. In the absence driven by a higher wage bill: wages in relation to of government, basic services such as health GDP rose significantly from 2005 to 2014 (from and education are mainly provided by non- 3.5 to about 6 percent of GDP). government actors. Seventy to eighty percent of primary school students attend non-public 5. Priority spending in Haiti has also increased. schools, placing a substantial financial burden on The fragmented fiscal data available would households and delivering achievements closely suggest that these additional resources have linked with household income. Outcomes are been channeled to the reconstruction and social equally unfavorable to the poor in the health sectors, consistent with the decline in poverty sector and only a small share of the poor can and improvements in human development benefit from basic safety nets. indicators (such as in education) observed over the same period. Overall, the share of resources 8. Haiti needs now to adjust to tighter budget allocated to social sectors seems to have constraints. An expansion of aid and expanded from an average of about 16 percent concessional borrowing has allowed Haiti to of total public spending for the period 2007 to increase public spending and catch up with 2010 to about 28 percent of total public spending regional comparators. It has, however, also after the earthquake, for the period 2010 to 2012. increased its reliance on foreign assistance In addition, 22 percent of donor assistance on as source of financing. Donor assistance and average is estimated to have gone towards social concessional financing represent about 70 sectors over the period of 2010 to 2012. percent of the financing for public capital spending. Social sectors including health, 6. The composition of public spending may, education and social protection rely on donor however, not be conducive to better services. assistance for 45 percent of their financing Part of the growing wage bill reflects, however, in 2012. This heavy reliance on donor and increases in support, administrative or security concessional financing makes these spending staff, and not in staff directly delivering social items particularly vulnerable to the decline in services. Furthermore, operating costs more xxvi BETTER SPENDING, BETTER SERVICES aid and in international oil prices (which affect Findings the availability of concessional resources from Venezuela). These tighter constraints could Greater and More Equitable Revenue put into question some of the recent progress Mobilization achieved in human development, making the country’s balancing act between developmental 10. While own fiscal revenue has increased over needs and fiscal sustainability even more the past decade, Haiti has still one of the lowest challenging. In this regard, this report examines revenue mobilization rates in the region. more particularly how to: Haiti’s own fiscal revenues have improved going from less than 10 percent of GDP in 2004 to i. mobilize greater fiscal revenue; 12.6 percent of GDP in 2014. Despite this rise enhance the growth dividend of public ii. in revenue, Haiti has still one of the lowest investment; rates of revenue mobilization in the region. iii. improve the efficiency of spending in critical Furthermore, Haiti’s tax system tends to be sectors such as health, education, and social regressive, relying heavily on indirect taxes. The protection; and ratio of direct to indirect taxes stood at about 30 percent in 2011, a level inferior to that of most iv. protect the budget from a return of fuel price countries in LAC and to the average of Low- subsidies. Income Countries, and is largely explained by 9. Against this backdrop, this study aims at the fact that a sizable share of Haiti’s revenues contributing to an evidence base to inform comes from international trade. Tax systems decision-making. This report builds on newly relying relatively more on direct taxes tend to available data, including the 2012 household be more progressive because in such systems the survey, the 2012 Demographic and Health burden of taxes weighs differently on economic Survey (DHS), and the 2011/12 School Census, agents of varying income levels. Moreover, while to provide a better understanding of the equity direct taxes tend to be harder to administer, they and sustainability issues in the delivery of some are frequently seen as an important indicator basic services such as health and education, and and tool of successful state-building as they tend engage policy makers and other actors on gaps to create a more direct citizen-state interaction and priorities. The surge in donor assistance and and voice. the availability of concessional financing have 11. With income tax rates largely comparable to exacerbated the fragmentation of fiscal data in the regional average, a large share of revenue Haiti. As part of the study, a special effort has is, however, lost because of exemptions. been made to partially consolidate fiscal data and Corporate and personal income tax rates are gather a more comprehensive picture of the use largely comparable to the regional average of public resources (BOOST Initiative). Finally, and cannot explain Haiti’s low tax burden. For the document builds on the recently completed instance, Haiti’s corporate income tax rate of 30 Public Investment Management Diagnostic and percent is slightly above the regional average. Assessment (PIMDA) that has led to an action Exemptions, however, are estimated to have plan to improve the management of public amounted in 2011 to the equivalent of more than investment in Haiti. half of all income tax collected. Most exemptions derive from the 2002 investment code that grants 15-year-tax exemptions for companies in free xxvii zones as well as 5-10 year exemptions for specific taxes. For tariffs, the revenue shortfall rises to 50 investment projects that are considered desirable percent and the classification of exemptions in from a development perspective. International customs data is insufficient to thoroughly track experience tends to show, however, that lower their justification. taxation does not compensate for an unfavorable 14. Finally, while Haiti is one of the most open investment climate and foster little additional economies, inspection fees increase the investment. Anecdotal evidence suggests a country’s effective protection. While Haiti is number of taxes that generate little revenue but one of the most open economies with very low impose high compliance costs on businesses tariffs, it imposes various fees that add up to the (“nuisance” taxes) and reducing them would be cost of importing goods. A 5 percent inspection a better alternative to attract investors. fee is applied to almost all imported products, 12. The top personal income tax rate seems to increasing tariff revenues by about 75 percent in apply only to very few taxpayers. While top 2013. Inspection fees in the magnitude applied income tax rates are also in line with other in Haiti mask the true level of protection in the countries in the region (30 percent), income country. Being applied across the board with an brackets are defined in such a way that the top identical rate for almost all products, they are rates only apply to a small share of taxpayers. unlikely to be in line with the country’s strategic There are currently four tax bands for personal trade objectives that would normally suggest income tax: 10 percent on incomes above HTG placing a higher tariff on consumer goods, and 20,000 and below 100,000; 15 percent up to lower tariffs on key inputs and capital goods. HTG 250,000; 25 percent up to HTG 750,000; Meanwhile, Haiti is a member of CARICOM, but and 30 percent for everything above that. The applies lower tariffs. Alignment of Haiti’s tariffs maximum rate is thus levied at incomes above to the current CARICOM Common External more than twenty times the national average per Tariff could generate significant additional capita GDP, more than twice the ratio for the revenue, but also substantially raise protection next highest comparator country (Guatemala at and costs for firms and consumers. One way about eight times). forward could be to replace the inspection fees with a combination of higher regular tariffs and 13. Substantial fiscal revenue shortfalls also occur a removal of exemptions with no loss in total at the borders. The largest share of fiscal revenue revenue and no increase in protection. in Haiti (60 percent) is collected at the border through customs duties, inspection fees, and sales and excise taxes on imported goods. While Getting a Greater Growth Dividend from no complete data is available on TCA and excise Public Investment tax exemptions, detailed transaction level data on taxes collected at the border makes it possible 15. Despite rising public investment, economic to calculate the value of exemptions on imported growth has not accelerated so far in Haiti. Haiti products. Estimating the foregone revenue by still performs poorly for selected infrastructure comparing the actual revenue collected to the indicators, such as access to electricity, roads or amount an importer should have normally paid ports. In this context, high public investment had the official rate been applied suggests that should be expected to contribute in reducing Haiti gave up in 2013 the equivalent of 14 percent bottlenecks to faster growth. This paradox of total sales tax revenue collected at the border, is, however, nothing new in Haiti. Several plus another 19 percent in ad valorem excise reasons have been put forward, ranging from xxviii BETTER SPENDING, BETTER SERVICES deficiencies in the country’s national accounting by the average LAC country. The gap in health system to chronic lack of maintenance or simply outcomes between rich and poor households the unproductive nature of the investment itself. has also narrowed. Significant gains have been Poor past donor coordination and the high made in infant mortality for the poorest quintile, volatility of external aid have also been argued to as well as the second and third quintiles over have affected the impact of investments in Haiti. the last decade. Reduction in stunting and in the prevalence of diarrhea has been more 16. This disappointing outcome stems in part from pronounced for poor income groups. very weak public investment management. Haiti’s public investment management exhibits 18. Health indicators in Haiti remain, nevertheless, a number of distinctive features and practices poor and disparities persist. The under-five common to countries that are aid-dependent, mortality rate is four times higher than the LAC including weak appraisal capacity and reliance regional average. The 2012 DHS generally shows on donors to design good projects. These hamper large inequalities in health outcomes and results the effectiveness of public investments. Sectoral that are worse for the poorest quintiles. Infant strategies to guide the prioritization of projects mortality was thus found to be 62 (per 1,000 live are lacking. This leads to a Public Investment births) for the highest quintile of income, versus Program composed of projects that are neither 104 for the lowest quintile of income in 2012. fully assessed nor prioritized. Furthermore, there Compared to the highest quintile, the number of is no effective ex-ante control on disbursements children suffering from stunting was four times based on the physical progress of projects higher in the lowest quintile in 2012 (DHS, against plans. More importantly, domestically- 2012), a gap that has not changed since 2006. funded capital expenditures are not properly 19. While the level of total health expenditure accounted for, tracked and reported, creating an per capita is relatively high in Haiti, public environment conducive to a lack of transparency spending in health is very low. Health and accountability, as well as to mismanagement expenditure per capita in Haiti has increased in of scarce public resources. Finally, even though line with the average of low-income countries the existing legal framework is acceptable and at USD 76.6 per capita stood in 2012 above for the management of public investments, what the average low-income country spends. its requirements are rarely respected, with According to the WHO, USD 50 per capita numerous processes and procedures that, when would ensure the provision of a basic care not redundant, are excessively elaborate. package to the population. Haiti was allocating, however, only 5 percent of government spending Increasing Sustainability and Improving to health in 2013-14, well behind the 15 percent Further Health Outcomes recommended by the Abuja Declaration, and government spending accounted for only 17. Health indicators have improved over the about 1.5 percent of GDP, a low rate compared last two decades. Child health indicators have to other low- and middle-income countries, improved for instance with the under-five spending on average between 2 and 6 percent of mortality rate declining from 145 deaths per GDP. This high level of per capita expenditure 1,000 live births in 1990 to 73 deaths per 1,000 was thus mainly financed by external funding, births in 2013. Haitians can now expect to live which surged after the earthquake but has been 8 more years than back in 1990, two additional declining since 2012. years gained compared to the progress made xxix 20. Despite the recent increase in donor assistance, 22. The productivity of the medical personnel lack of finance still prevents the poor from of first-level primary facilities is low. Only seeking treatment. In 2012, total health 13 percent of health facilities are estimated expenditure was largely financed by donors to produce health services efficiently: they in Haiti, more so than in other low-income have neither excessive operating costs, nor countries, questioning the sustainability of the excessive personnel. In particular, the number recently-achieved progress. The surge in donor of consultations is low with one patient per assistance seems to have been accompanied hour, much lower than in comparable countries. by a substantial decline in the contribution of A second occupation could explain low households to their health expenditure. This productivity with evidence revealing often that development should have improved access to half of the medical personnel derive an income health services for the poorest. Yet, the most from sales activities or from a private practice. recent household survey reveals that 65 percent Excessive administrative personnel may also of households in the lowest consumption undermine efficiency. Administrative personnel quintiles did not seek medical attention due to represents 40 percent of total institutional their lack of money against 39 percent for the personnel although in general the need for highest consumption quintiles. Moreover, lack of such personnel in primary level facilities is low. money for treatment is most frequently cited as These facilities are small and have little or no a barrier to health care in Haiti by women aged equipment (for dispensaries), requiring a low 15-49 (76 percent). level of maintenance and management, and therefore few administrative personnel. 21. The composition of public spending may not be conducive to better service delivery. The health sector is a good example, showing that Keeping Children in School and Improving overall operational spending has not kept up Education Outcomes with the recent rise in public investment. While operating expenditure has changed little, the 23. Today, the majority of children are in school investment budget from public funds (Treasury in Haiti. The majority of preschool age children, and Petrocaribe) in this sector has doubled and 90 percent of children of official primary between 2006 and 2013. Moreover, the wage school age (6-11), are in school. Educational bill may be crowding out critical operating attainment has also been rising steadily across expenses (such as medicine and equipment) cohorts of young adults. In 1994, fewer than with the share of the operating budget allocated 30 percent of 15-19 year olds had reached lower to personnel costs representing about 90 percent secondary school, a figure which had risen to over the period 2006 to 2012. Furthermore, over 50 percent for women and over 40 percent public investment may not be targeted where it for men by 2012. is most needed: in 2012, for instance, nearly a 24. Despite this progress, education indicators third of the commitments of national investment remain low with late primary school entry, funds were allocated to the construction or high dropout, and limited learning. Compared rehabilitation of hospitals and only 22 percent of to its LAC neighbors, Haiti has the highest share the commitments were allocated to investment of adults with no education, and the highest projects for dispensaries and health centers, share of 15-19 year olds who have not completed although these facilities account for 87 percent primary school. The adult literacy rate is about of health care delivery in Haiti. 77 percent, midway between the LIC and the xxx BETTER SPENDING, BETTER SERVICES LAC averages. Furthermore, many students, the operating budget is allocated to salaries and particularly in poor communities, learn little. personnel costs. Although, this share is in line Indicators of the quality of education, including with other LAC countries (where 60-80 percent teacher knowledge and learning materials of the total education budget is absorbed by the available at the school level, suggest that many wage bill), the public sector in these countries children, particularly poor ones, are receiving pays the majority of all primary and secondary a low-quality primary education. Assessments teachers, while in Haiti the State only directly administered in early grades in selected schools employs those in the small public sector. As a have found that fundamental skills are acquired result, about 42 percent of salary expenditures very slowly or not at all, particularly in schools go to staff who are not working as teachers, in poor communities. For example, the average directors, or inspectors. Furthermore, half of third grader reads only 23 words per minute, non-salary operating expenditures go towards well below the estimated speed of 35-60 words textbooks, national exams, and the national per minute required for comprehension of a school canteen program. basic text. 27. Despite weak progression and learning, 25. Growth in education supply has primarily households spend a substantial amount come from the non-public sector, making on sending children to school. Among all Haiti one of the countries with the largest non- households with children age 6-14 in school, 93 public sectors at the primary and secondary percent report positive education expenditures. levels. According to the 2010-11 School These expenditures are substantial: on average, Census (the most recent published), 94 percent these households reported spending 10 percent of preschool, 78 percent of primary, and 73 of total annual household consumption percent of secondary students attend non-public on education in the 2011-12 school year. schools. These shares are among the highest in Consequently, costs remain a major obstacle the world. Tuition waivers for students in non- to education for poor households: when asked public primary schools have, however, been put why school-aged children are not currently in in place to help children access school. Donors school, 83 percent of households cite costs as the are financing a complete cycle of primary primary reason. school for cohorts of children in disadvantaged communities by providing tuition waivers to Towards a More Effective and Better Targeted non-public schools, while the authorities are Social Protection System financing a similar program financed by a new tax on international phone calls and money 28. The chronic and transient poor should be transfers. Declining donor financing and a recent an important focus of an effective social decision by the Ministry of Education to stop protection system in Haiti. Although poverty funding tuition waivers for new cohorts of first is widespread in Haiti, rural areas and specific graders in non-public schools could therefore geographic departments require special attention threaten the gains in access made in recent years. from the social protection system. Nearly half 26. As in health, however, the composition of the households are considered chronically poor public spending is not conducive to better because they live below the moderate poverty services. The operating budget of the Ministry line and lack at least three of the seven basic of Education is skewed towards non-“front line” dimensions defining non-monetary well-being. staff. In 2013, for instance, almost 70 percent of Such a large share of chronic poverty, unusual xxxi in Latin America, highlights the structural network of multi-sectorial agents, as well as local challenges of poverty reduction and limitations coordination on social protection. of the contributions of social protection 30. Despite these efforts, social protection interventions without combined interventions spending has still a long way to go to become around basic services and income generation. more aligned with challenges at hand. More Haiti’s social protection system should also investment is particularly needed for children be able to include the transient poor with one under five as this group suffers from the highest million people living slightly above the poverty poverty rates and as a lack of investment in young line but vulnerable to drop below if a shock children has irreversible negative consequences. (weather or health-related) occurred. Natural As shown in the education and health findings, disasters, in particular, have great disruptive there is a need for a set of social protection and potential partly because they have such a serious promotion interventions that would enable the impact on agriculture, which represents the poorest households (especially those in rural main source of livelihood for most of the poor, areas with young children) to overcome hurdles especially in rural areas. in building, accumulating, and preserving their 29. During the last three years (2012-14), there human capital in the face of repeated shocks. have been encouraging signs of increased 31. Furthermore, the various programs attention for social protection as a framework building Haiti’s social protection system to address both short-term and long-term are excessively fragmented, small, and not vulnerability and poverty reduction. In 2012, targeted well enough to make a difference. the national strategy for fighting hunger and Within government’s public spending on malnutrition, ‘Aba Grangou’, was launched, with social protection, both the contributory (social the aim of coordinating donor and government insurance) and non-contributory pillars (social activities in the nutrition sector. A notable step safety nets) are spread across multiple institutions, toward the formation of a social protection challenging the country’s weak administrative strategy was taken in 2014 with the launch of the capacity. Public spending in social protection is three-year action plan “Thinking and Fighting low, amounting to about 5 percent of GDP with for a Haiti without Poverty: Action Plan for subsidies for electricity and petroleum products Accelerating the Reduction of Extreme Poverty” representing a considerable share (2 percent of (PAARP). This document defines a set of flagship GDP). In addition, social assistance benefits are programs constituting a social assistance strategy not well targeted: the results of the most recent named “EDE PEP” to protect the vulnerable household survey suggest that as much as half of population living in extreme poverty throughout social assistance benefits accrues to the non-poor. their life cycle, in the short and medium term, to As a consequence, social protection programs ensure long-term investment in human capital have limited impact on poverty and inequality. and to provide opportunities to overcome the Finally, most EDE PEP programs are financed condition of extreme poverty. The PAARP also principally by concessional borrowing, which identifies a set of elements required to underpin poses challenges in terms of their sustainability. the implementation of a social protection system, such as a national targeting system, a unique Preventing a Return of Fuel Price Subsidies beneficiary registry, and an integrated service delivery model at communal level through a 32. The size of oil subsidies in Haiti had increased dramatically over the last five years, imposing xxxii BETTER SPENDING, BETTER SERVICES a large burden on public finances. Haiti raised benefits from a universal reduction in education, retail fuel prices in October 2014, a notoriously health, and food prices. difficult reform, especially in a fragile situation. 34. Increases in fuel prices have large effects on This move combined with a decline in electricity and transportation, however, and international oil prices has allowed to eliminate poor households are especially vulnerable to the fuel price subsidies that were burdening the increases in food prices. Directly, households budget. Like other net oil-importing developing are affected through their own consumption of countries, Haiti has faced economic difficulties gasoline or kerosene. Indirectly, they are also with high oil prices. Haiti has been for a long affected since petroleum products are used as time cautious passing through the increase in intermediary products in many sectors and their international oil prices to consumers. In 2010, higher price will feed into the price of the final it abandoned a mechanism of automatic pass- good produced by these sectors. The analysis of through, and adopted a price control system the indirect effects suggests that an increase in through adjustments in taxes. As the gap between fuel prices would have the largest direct impact international and domestic prices widened, the on transportation and electricity, in which oil cost of keeping retail petroleum prices fixed products account for more than 30 percent of was placing an increasingly significant burden total inputs. Even though the production of on public finances (nearly 2 percent of GDP in food is not intensive in fuel, rises in food prices 2013). nevertheless impose a sizeable loss of welfare on 33. Subsidized fuels are mainly consumed by the the poorest households, due to the large share of wealthiest households. Poor households in food in their budgets. Haiti consume very little fuel, both in absolute 35. The budget remains vulnerable to a rebound terms and as a share of their total budget. As a in international oil prices. Fuel subsidies have result, the poor receive only 1.6 percent of total been eliminated in October 2014. International subsidies accruing directly to households. Since oil prices are, however, volatile and without the direct usage of petroleum products is so limited introduction of an automatic price adjustment among the poor, wealthier households benefit mechanism for petroleum products, fuel price disproportionally from subsidies granted to subsidies could return to haunt the budget, but these products: over 93 percent of fuel subsidies this time under tighter financing constraints. going directly to households benefit the richest The subsidies, which amounted to around 2 20 percent of the population. Even though percent of GDP in 2013, could climb to over 3 there are no data on expenditures on kerosene, percent of GDP with international prices at USD there is some indication, nevertheless, that poor 100/barrel. households in rural areas would be the most affected by increases in the prices of this product: Going Forward among the rural poor, over 80 percent of households use kerosene as main source of fuel. 36. Against this background, some options going Fuel subsidies are particularly regressive when forward present themselves. While the report compared to subsidies on other categories of wanted to provide mainly a diagnostic and a consumer expenditures, such as food, education better understanding of some critical sectors and health: because these items represent a such as health and education, some broad larger share in their budget, the poor would options going forward can nevertheless be receive between 30 percent and 40 percent of the suggested as basis for discussion: xxxiii • Data: Overall tighter financing constraints and maintenance expenses (e.g., equipment, make it even more urgent to collect timely utilities), thereby limiting growth dividends data to track public spending appropriately of investment. and make it more efficient in different sectors. • Fiscal revenue: Balancing fiscal sustainability Differences in budget classifications and and development needs will remain a the absence of an effective Single Treasury daunting challenge for Haiti as long as the Account lead to a fragmentation of fiscal data country does not mobilize substantially more and prevents a comprehensive monitoring of own fiscal resources. In this regard, a review public spending from budget appropriation would be needed of existing exemption to payment. Fiscal reporting needs thus to be regimes for all tax instruments with priority strengthened to allow an appropriate tracking to customs duties and corporate income tax of public spending and set in place the where revenue losses are the greatest. An minimum conditions for transparency and important step would be the introduction of a accountability in the use of public resources. clear classification system in customs data to Completing the roll-out of the Single track the justification of exemptions. Where Treasury Account and unifying the budget tax exemptions or other investment incentives classifications among the various sources of are used, a clear framework needs to be put financing would be important steps in this in place to measure and track their costs. At direction. the same time, it is important to track and • Donors: With declining donor assistance, monitor the benefits these incentives were greater donor coordination is needed to make expected to deliver (such as employment the most of these resources. A master plan creation or technology transfer). for critical sectors translating the priorities of • Greater growth dividend: Haiti’s these sectors into areas of intervention could development needs are daunting and will federate the actions of the donor community, need to be addressed. In this regard, the allowing it to coordinate its technical and overall public investment management needs financial contributions, and avoid potential to be improved through a strengthening of duplication and inefficiencies. Such a project evaluation, selection, programming, plan would have indicators and deadlines, execution, and control. As a first step, the triennial operational plans, and an annual priority should be put on implementing action plans to structure the monitoring of its basic elements of formal project selection implementation. and appraisal, execution and controls such • Budget composition: The composition of as the need for investment projects to (i) public spending has room for improvement advance only if they are mature enough (i.e. to enable better service delivery. The growing developed in a concept or in a justification); wage bill, although in line with economies at (ii) be monitored (including physically); (iii) similar levels of development, has been the use resources adequately; (iv) are controlled driving force of total current spending, but ex-post as required. most of it is accounted for by staff not directly • Improve the performance of health facilities: involved in service delivery. Sharp increases For a long time the focus has been put on in capital outlays have not contributed to increasing the resources or the inputs. With the acceleration of growth, due in part to an shrinking budgets, the time may have come inadequate resource allocation for operating to shift the logic upside down and pay more xxxiv BETTER SPENDING, BETTER SERVICES attention to the results or the services actually achievements. Greater oversight could also delivered by the health facilities through a be called for through a mandatory teaching more general use of results-based financing license based on demonstrated competencies (RBF). RBF could provide monetary and and a mandatory school identity card, leading non-monetary incentives to health workers to certification. In addition, targeted in- in order to increase their productivity, service training programs for teachers could discouraging them from exercising another be put in place to improve instruction in the occupation. Health facilities may also have classroom and learning. very low productivity because they have too • Better targeting in social protection: Haiti’s few patients. To improve the use of their social protection suffers from fragmentation resources, as well as health outcomes, such of programs and interventions. The tighter facilities may want to reinforce primary health resource constraints related to a reduced level services and operate mobile clinics to access to concessional borrowing and donor increase access to health services, particularly assistance provide an opportunity to design a for the poor. less fragmented and better targeted system of • More accountable schools to increase social safety nets. In the meantime, continuing learning in the classroom: The majority of progress in the implementation of the social schools at all levels in Haiti are non-public, and registry of potentially eligible beneficiaries is operate with little oversight or accountability. required. With limited public resources, the sector will • Protecting the budget: With fiscal resources remain predominantly non-public for the remaining limited, unbounded calls on the foreseeable future and efforts should focus on budget should be contained. In this regard, making the schools more accountable for the the re-emergence of fuel price subsidies quality of service they deliver. In this regard, remain a risk. Higher international oil prices efforts could focus on the comprehensive could come back again to haunt the budget. identification of schools or students benefiting There is thus a need to return to an automatic from the different programs offered by the pass-through with this mechanism given to Ministry of Education, including the donor- an independent agency to reduce political financed Tuition Waiver Program (TWP) and influence in the setting of fuel prices. PSUGO and linking this support with school xxxv MAP xxxvi BETTER SPENDING, BETTER SERVICES Chapter 1 : Country Profile This chapter reviews briefly Haiti’s economic potential and discusses the main obstacles the country is facing in achieving faster economic growth. Haiti has comparative advantages, including its proximity and access to major markets, a young labor force and a dynamic diaspora, and substantial geographic, historical, and cultural assets. A history of political instability, violence, and natural disasters, as well as poor governance, inadequate infrastructure, and limited skills, has prevented the country from realizing up to now its aspirations. A. Haiti Has Opportunities …. Areas of economic opportunity for Haiti lie in agribusiness, light manufacturing and tourism. 1.1. Haiti’s geography, resources, and history According to the World Economic Forum, Haiti’s provide it with opportunities. Haiti occupies economic fundamentals could allow the country the western, smaller portion of the island of to become a vibrant economy and grow by 6-8 Hispaniola, which it shares with the Dominican percent a year if adequate policies were in place.2 Republic. Both by area and population, Haiti is the third largest Caribbean nation (after Cuba market access to the United States and Canada, but and the Dominican Republic), with 27,750 has not succeeded in fully exploiting this advantage. square kilometers and an estimated population For example, the country has not been able to fill of 10.4 million people. The country has its quota for exports to the United States under the Hemispheric Opportunities through Partnership comparative advantages, including its proximity Encouragement II (HOPE II) initiative or under the and access to major markets; a young labor Haiti Economic Lift Program (HELP) Act. These laws force and a dynamic diaspora; and substantial allow Haiti to assemble textiles, whatever the origin of geographic, historical, and cultural assets.1 the imported fabrics, and to export them to the United States duty free and tax free but are due to end in 2020, unless renewed. 1 Haiti has traditionally enjoyed substantial preferential 2 World Economic Forum (2011). 1 B. … Hampered by Political 2015, leaving the executive to broker a deal with the opposition and agree on a transitional Instability and Natural Disasters government that would organize elections. 1.2. Overall Haiti’s growth performance over the 1.5. Furthermore, the Haitian population is last four decades has been disappointing and one of the most exposed in the world to poverty remains endemic. A history of political natural disasters—hurricanes, floods and instability, violence, and natural disasters earthquakes.3 Between 1971 and 2013, Haiti’s has prevented the country from realizing its economy has been subjected to numerous aspirations, trapping the country in a low shocks with natural disasters occurring almost equilibrium and keeping it as one of the poorest every year and adverse effects on economic and most unequal equal countries in the world. growth (Figure 1.2). The country has a higher GDP per capita fell by 0.7 percent per year on number of disasters per km2 than the average average between 1971 and 2013. As a result, the of the Caribbean countries (Table 1.1). In 2008, overall poverty headcount amounts to about 59 tropical storms and hurricanes caused losses percent and extreme poverty to 24 percent in estimated at 15 percent of GDP. The earthquake 2012, indicating that almost 6.3 million Haitians on January 12, 2010 killed 220,000 people, cannot meet their basic needs and 2.5 million displaced 1.5 million people, and implied losses cannot even cover their food needs. of the equivalent of 120 percent of GDP. 1.3. Political violence has occurred regularly throughout Haiti’s history, leading to C. … As Well as Weak Structural government instability (Figure 1.1). At Policies and Institutions Independence in 1804, Haiti was at the forefront of history, being the first nation to abolish slavery. 1.6. Haiti’s weak structural policies and institutions Since then, however, with the exception of the hamper faster economic growth (Figure 1.3). 30-year period of autocratic rule under Francois Poor governance and political instability have Duvalier (Papa Doc) and his son Jean-Claude been identified in previous assessments as the Duvalier (Baby Doc) (1957-1986), Haiti has major impediments to sustainable development, known a succession of short-lived governments. along with weak public sector capacity and Lacking sufficiently long periods of stability, the accountability, followed by low levels of country has struggled to develop the institutional education and badly deteriorated infrastructure.4 mechanisms and policy fundamentals essential Doing Business Indicators 2015 rank Haiti at 180 to development progress. out of 189 countries, while the 2014–15 Global 1.4. The past decade has been comparatively stable, Competitiveness Index (GCI) ranks it 137 out nevertheless. Over the past ten years, with of 144 countries. According to these indicators security provided by a large United Nations and the investors’ survey of the GCI, the major Stabilization Mission in Haiti (MINUSTAH), constraints in Haiti’s business environment two presidents have been chosen by election and include poor infrastructure, limited access the current President, Michel Martelly, was the to finance, and cumbersome administrative first to accede to the Presidency by election from procedures to start a business. a party in opposition. However, stability remains 3 From 1993 to 2012, Haiti has experienced two fragile, with frequent changes in Government droughts, one earthquake, 31 floods and 26 tropical and repeated delays in electoral calendars. Such storms/hurricanes. delays led to the lapsing of Parliament in January 4 See for instance World Bank (2002). 2 BETTER SPENDING, BETTER SERVICES Figure 1.1 : Annual GDP Growth vs. Change in Figure 1.2 : Annual GDP Growth vs. Occurrence of Government, 1971-2013 Natural Disaster, 1971-2013 (percent) (percent) 10 10 5 5 0 0 -5 -5 -10 -10 -15 -15 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 Occurrence GDP growth Occurrence GDP growth Sources : UN, World Bank and International Media Sources : UN, World Bank and EM-DAT: The OFDA/CRED International Disaster Database Table 1.1 : Frequency and Impact of Natural Disasters, 1971-2014 Country/ Number Disasters/ Disasters/ Disasters/ Deaths/ Total Damage/ Group of Natural Year Land Surface Population Population GDP Disasters (`000 sq. km) (millions) (millions) Haiti 137 3.1 5.0 13 23,427 1.776 exc. Earth-quake (2010) 136 3.1 4.9 13 1855 0.22 Dominican Rep. 60 1.4 1.2 6 311 0.05 Jamaica 34 0.8 3.1 13 102 - Nicaragua 66 1.5 0.5 11 2363 0.33 Honduras 70 1.6 0.6 9 3298 0.40 El Salvador 51 1.2 2.5 8 687 0.34 Guatemala 82 1.9 0.8 5 1754 0.12 Costa Rica 58 1.3 1.1 12 72 0.04 Panama 46 1.0 0.6 12 80 0.01 Other Caribbean States* 129 2.9 0.3 30 86 0.19 Source: EM-DAT: The OFDA/CRED International Disaster Database. *Antigua and Barbuda, The Bahamas, Barbados, Belize, Guyana, Suriname, Dominica, Grenada, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines and Trinidad and Tobago. Simple average. 3 Figure 1.3 : Obstacles to Growth Governance Is a Concern... … Infrastructure Inappropriate, ... Governance Indicators , 2013 Quality of Port Infrastructure, 2013 (percentile rank) (1=extremely underdeveloped to 7=well developed and ef cient by international standards) Control of Corruption Low Income Countries Rule of Law Latin America and Caribbean Government E ectiveness Small Caribbean Countries Voice and Accountability Haiti 0 50 100 0 2 4 6 Haiti LIC LAC Source : World Bank Source : World Economic Forum ... Especially Access to Electricity, ... … Human Development Low… Access to Electricity, 2011 Human Development Index, 2013 (percentage of population) (overall score) Least Developed Low Income Countries Countries Caribbean Latin America and Caribbean Latin America and Caribbean Haiti Haiti 0 50 100 0 0.2 0.4 0.6 0.8 Source: World Bank Source : UNDP 4 BETTER SPENDING, BETTER SERVICES Figure 1.3: Obstacles to growth (continued) … And the Business Climate Unfavorable … … Restricting Competition and Opportunities … Business Indicators, 2015 or Latest Intensity of Local Competition, 2015 (Doing Business) (1=low to 7= high, scores and position of Haiti) 49 Private credit (In % of GDP, 2013) 31 Latin America 19 and Caribbean Strengh of minority 5 inverstor protection 5 (0-10) 2 Low Income Countries 31 Starting a business (days) 29 97 Haiti 140/144 0 50 100 150 LAC LIC Haiti 3.0 3.5 4.0 4,5 Source : World Bank Source : World Economic Forum ... And Encouraging the Population to Seek ... Or in Low Paid Jobs Its Future Abroad … Underemployment, 2012 Stock of Migrants of Hatian Origin, 1960-2010 (percentage of employed population ) (in thousands) 1200 Sud-Est 79.7 Nippes 79.5 1000 Grand'Anse 79.3 Nord-Ouest 78.2 Sud 77.7 800 Centre 77.1 Nord 76.0 600 Artibonite 75.5 Nord-Est 74.6 400 Ouest 56.4 200 National 70.0 Urban 57.3 Rural 80.3 0 1960 1970 1980 1990 2000 2010 0 20 40 60 80 100 Source : World Bank Source : ECVMAS 5 behind its competitors (Figure 1.5).5 The road Figure 1.4 : Change in Governance Indicators, network in Haiti is in poor condition and many 2004-13 (Percentile Rank) parts of the country have little connectivity.6 Haiti is also less integrated into the global shipping Control of Corruption line network than many developing countries.7 Furthermore, the costs of loading and unloading Rule of Law a standard container at Port-au-Prince are by far the highest of the Caribbean ports (Figure 1.6).8 Government E ectiveness 1.9. In particular, the quality of public utilities is Voice and poor. The provision of electricity is unstable Accountability and there are frequent power cuts and surges, 0 5 10 15 20 which can result in serious damage to industrial Source : World Bank equipment. Despite this poor quality of service, the cost of electricity is among the highest in the region. In 2011, industry was charged 1.7. Although governance indicators have improved, they remain low. The increase in 5 The LPI is based on six core dimensions of trade- political stability after the low point of 2004 related services, including customs, infrastructure, international shipments, logistics competence, came with an improvement in a number tracking and tracing, and timeliness. of governance indicators (Figure 1.4). The 6 The sharp deterioration of Haiti’s road network, the implementation of the Automated System for inadequate packaging of fruit and vegetables and the Customs Data (ASCUYDA) was stepped up, use of vehicles unsuited to the transport of agricultural for instance, after the earthquake. Indicators for goods are responsible for considerable commercial control of corruption, rule of law, government losses. In the North West, for instance, transport costs for a stem of bananas from the farm to the primary effectiveness, as well as voice and accountability market are estimated to represent 25 percent of the have all registered improvements in Haiti over sale price or 45 percent of the profit margin (World that period, although remaining low compared Bank, 2013). to the average of the LAC region and to that of 7 According to UNCTAD’s liner shipping connectivity low-income economies. index (LSCI), Haiti ranks lower than the main players in the Caribbean: Haiti’s score of 5 is much lower 1.8. Haiti’s infrastructure also falls short. Island than scores by Jamaica (21), Bahamas (27), and the economies are extremely dependent on the Dominican Republic (23) that host transshipments quality, frequency and cost of the means of ports. The liner shipping connectivity index (LSCI) transport that link them to markets which is made up of five components: 1) the number of companies that provide services from/to a country’s represent both outlets for their products and ports; 2) the size of the largest ship providing services supply sources for the needed imported goods. from/to a country’s port (measured in Twenty foot The efficiency and effectiveness of transport, Equivalent Units--TEU); 3) the number of services whether by road, by sea, or by air, therefore that connect the country’s port(s) to other countries’ contribute to the competitiveness of these ports; 4) the total number of ships operating from/to countries. The quality of transport and logistics the country’s port; and 5) the total container carrying capacity of those ships. services in Haiti is poor, however, ranking 144th 8 According to a 2009 study mandated by the Haitian out of 160 countries on the World Bank’s logistics Chamber of Commerce, costs in Port–au-Prince are performance index (LPI) in 2013 and trailing five times higher than in the ports of the Dominican Republic (World Bank, 2013). 6 BETTER SPENDING, BETTER SERVICES Figure 1.5 : Logistic Performance Index, 2014 (1=lowest, 5=highest) 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Overall score Customs Infrastructure International shipments Logistics quality Tracking and tracing Timeliness Bahamas Dominican Rep. Jamaica Guyana Haiti Source : World Bank Figure 1.6 : Port Tari s Estimated Cost Per TEU*, 2009 (in US Dollar). Port Everglades Miami Port of Spain, T&T Port of Point Lisas, T&T Dominican Ports Authority , T&T Puerto Rico Ports Authority Haiti Terminal Varreux Private Operators (Public Berths) 0 200 400 600 Stevedoring Wharfage Ship Charges** Source : TranSystems *Based on a 700 TEU ship, discharging and loading a mixture of 20-ft and 40ft containers. ** Pilotage, Port Dues, Light Dues & Dockage $0.32/KwH, compared to $0.18 in Nicaragua, percent of the level in the Dominican Republic $0.17 in the Dominican Republic and $0.06 (Figure 1.7). Access to water and sanitation in in Bangladesh.9 Furthermore, the supply of Haiti remains also limited.10 electricity covers only a small proportion of 1.10. Surveys also point to the lack of qualified the country, with rural areas being particularly labor.11 One major constraint faced by neglected. Per capita consumption of electricity in Haiti is substantially lower than in other 10 64 percent of the population has access to an improved Caribbean countries, for example, it is only two drinking water source (48 percent in rural areas) and 31 percent access to improved sanitation facilities (16 percent in rural areas). 9 IFC (2011) 11 See for instance ILO (2010) 7 Figure 1.7 : Electric Power Consumption, 2011 (kWh Per Capita). 900 800 700 600 500 400 300 200 100 0 Dominican Rep. El Salvador Honduras Bolivia Guatemala Nicaragua Haiti Source : World Bank enterprises in Haiti is the difficulty in finding technicians that are well qualified, particularly in new technologies. This forces the country to position itself as a low cost producer for goods and services requiring limited skills. A deficient labor market may be at fault, with no institutional mechanism to enable the exchange of information between labor demand and supply. More generally, the level of education among the adult population remains low: 45.7 percent of the adult population (60.5 percent of households’ heads) have never attended school or have not completed primary education. Not mastering basic skills such as literacy and numeracy when starting work represents a major impediment for their insertion in the labor market and, more importantly, for their ability to absorb post-school training either on or off the job, and to adapt to changing job requirements. 8 BETTER SPENDING, BETTER SERVICES Chapter 2 : Living With Tighter Budget Constraints12 This chapter discusses the overall trends in public expenditure, fiscal revenue, and borrowing. Haiti has managed over the past decade to increase its own fiscal revenue. The re-engagement of donors led to an increase in assistance that accelerated following the 2010 earthquake. Debt relief has also provided the country with space to access concessional borrowing. The availability of these resources has allowed public spending in Haiti to catch up with comparators. The sustainability of this high level of public spending is, however, now put into question by the decline in aid and lower oil international prices that are reducing the availability of concessional financing. Against this backdrop, the need to mobilize own fiscal revenue and improve the efficiency of public spending have become all the more urgent. A. Macroeconomic Stability Has Been digits and has remained contained at these levels since 2009 (Figure 2.1). International reserves Maintained …12 have increased and have been maintained at 2.1. Haiti’s macroeconomic environment has an adequate level, covering about 5 months of broadly improved over the past decade. Growth imports (Figure 2.2). Between 2009 and 2011, has regained some momentum following the Haiti’s stock of external debt drastically shrank earthquake, driven by services and construction following debt relief. In addition to benefiting (Table 2.1). Inflation was brought back to single from the HIPC and the MDRI initiatives, Haiti received additional debt cancellation in the aftermath of the 2010 earthquake. As a result, 12 Prepared by Kassia Belo da Silva Antoine, Ibrahim El ghandour, Michel Matera, Emilio Sacerdoti, and total external debt fell to 8.9 percent of GDP in Rafael Van der Borght. 2011. 9 Figure 2.1 : In ation, 2004-14 Figure 2.2 : Gross International Reserves, 2004-14 (In Percentage) (Millions of US Dollars) 30 30 25 25 20 20 15 15 10 10 5 5 0 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Sources : IMF and Haiti’s Staticals Institute (IHSI) Source : Central Bank of Haiti (BRH) B. … With Greater Revenue the Government of Haiti, these resources have nevertheless contributed to the recent increase Mobilization and a Surge in Aid in fiscal revenue (Figure 2.3). 2.2. Haiti received an unprecedented amount 2.3. Domestic revenue mobilization has also of support and aid in response to the 2010 improved. Although more modestly, Haiti’s earthquake. With the re-engagement of donors own fiscal revenues have also improved during after the political turmoil, external grants these years going from less than 10 percent of increased from 2 percent of GDP in 2004 to GDP in 2004 to 12.6 percent of GDP in 2014 8.1 percent in 2013. Following the earthquake, (Table 2.1). Since 2011, however, progress in private donations are reported to have reached increasing revenue collection has slowed and USD 3.1 billion (47 percent of GDP). 58 donors even reversed with tax revenues in terms of GDP made pledges totaling USD 5.5 billion (83 declining from their peak in 2012 of about 13 percent of GDP) to help Haiti at the International percent of GDP. As discussed in Chapter 4, some Donor’s Conference in New York on 31 March, of the factors behind this decline relate to the 2010. The Office of the Special Envoy for Haiti inefficiency of the tax collection agencies, as well reports that bilateral and multilateral donors had as revenue shortfalls stemming from the freeze pledged USD 9.28 billion (140 percent of GDP) of domestic fuel prices and exemptions granted in humanitarian and recovery funding for 2010 on imports. to 2012. Of these pledges, USD 5.63 billion (60.7 2.4. This expansion in overall revenue allowed percent) has been disbursed. From the available Haiti to catch up with comparator countries. figures, it appears that NGOs and private Haiti’s overall fiscal revenues registered a gradual contractors have been the primary intermediate increase, reaching about 20 percent of GDP in recipients of this assistance, extending their 2014 (compared to 11.3 percent of GDP in already significant presence in the provision of 2004) and catching up with comparators (Figure social services (Ramachandran and Walz, 2012). 2.4). Haiti’s mobilized revenue is now closer Although very little money has gone directly to 10 BETTER SPENDING, BETTER SERVICES Table 2.1 : Macroeconomic Indicators 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 (annual percentage change) National account and prices Real GDP 1.8 2.3 3.3 0.8 3.1 -5.5 5.5 2.9 4.2 2.8 Consumer price index (average) 28.3 14.2 9.0 14.4 3.4 4.1 7.4 6.8 6.8 3.9 External sector Exports (goods, f.o.b.) 21.3 7.7 5.5 -6.0 12.4 2.2 36.3 2.2 12.3 3.4 Import (goods, f.o.b.) 8.0 18.3 4.5 30.2 -3.6 38.3 7.3 -11.1 24.3 1.9 Money and credit Base money (currency in circulation and 0.6 5.5 7.6 13.9 9.5 31.2 6.0 -3.7 15.1 1.7 gourde deposits) Credit to private sector 21.1 5.5 10.3 25.8 14.7 -6.3 25.5 30.6 16.8 11.2 (in percentage of GDP) Fiscal accounts Revenues 13.1 13.5 15.8 15.1 17.8 23.9 21.9 23.4 20.8 19.7 Expenditures 13.8 15.2 15.6 18.2 22.4 21.7 25.5 28.2 28.0 26.1 Overall balance -0.6 -1.7 0.2 -3.1 -4.6 2.2 -3.6 -4.8 -7.1 -6.4 Balance exc. grants -4.1 -5.0 -5.1 -7.5 -11.3 -10.0 -12.7 -15.4 -15.2 -13.5 Public. debt, Stock 34.2 28.2 25.5 30.0 19.4 15.9 11.5 15.8 19.5 24.1 Balance of payment Current account balance 0.7 -1.8 -1.5 -3.1 -1.9 -1.5 -4.3 -5.7 -6.3 -6.3 Exports (goods and services) 14.0 14.5 13.2 14.0 15.7 15.3 17.5 16.8 18.6 19.0 Imports (goods and services) 41.7 45.3 40.8 43.9 42.9 64.7 59.0 53.2 52.3 51.8 Workers remittance - - - 20.9 20.9 22.3 20.6 20.4 21.1 22.7 Foreing direct investment 0.6 3.4 1.3 0.5 0.8 2.7 1.6 2.0 1.9 1.1 External debt, Stock 34.2 28.2 25.5 30.0 19.4 12.9 8.9 13.8 17.7 21.4 Memorandum items: Nominal GDP (millions of gourdes) 168,035 197,183 220,110 250,590 267,880 266,952 302,854 328,061 364,526 388,809 Nominal GDP (millions of US dollars) 4,310 4,757 5,885 6,549 6,585 6,623 7,517 7,890 8,453 8,713 Exchange rate (average) 39.0 41.4 37.4 38.3 40.7 40.3 40.3 41.6 43.1 44.6 Gross reserves (millions of US dollars) 279 388 596 759 1,018 1,832 2,045 2,346 2,384 1,914 in months of imports 1.6 2.0 2.5 3.2 2.8 5.0 5.8 6.4 6.3 5.2 Sources : Ministry of Finance (MEF), Central Bank (BRH), Haiti’s Statical Institute (IHSI), IMF and BOOST database. * Note : Annual data covers Haiti’s fiscal year (Oct. 1 to Sep. 30) 11 Figure 2.3 : Fiscal Revenues, 2004-13 Figure 2.4 : General Government – Total Revenues, (In Percentage of GDP) 2000-12 (In Percentage of GDP) 25 40 34.1 34.0 35 29.2 28.9 20 26.4 26.7 26.3 25.9 30 23.1 22.6 20.1 15 25 17.0 20 10 15 10 5 5 0 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Haiti LIC LAC Island States Islands- LAC Fragile States Domestic revenues Customs Other revenues Grants 2000-2007 2010-2012 Source: Ministry of Finance (MEF) and IMF Source : IMF Table 2.2 : Fiscal Revenues - International Comparison (In Percentage of GDP) Small Caribb. Dominicain El Salvador Guatemala Costa Rica Nicaragua Honduras Jamaica Panama States Haiti Rep. LAC 2013 2013 2010 2013 2010 2012 2012 2012 2012 Total 20.8 27.5 13.6 18.4 17.4 18.9 11.6 13.9 18.7 26.4 23.1 Domestic 12.7 26.9 13.5 16.5 16 18.2 11.6 13.9 18.6 25.2 21.9 Grants 8.1 0.6 0.1 1.9 1.4 0.7 0 0 0.1 1.2 1 Source : MEF and IMF to the regional average and above the average C. … Allowing Higher Public Spending of countries at similar levels of development. The country is, however, much more reliant 2.5. Public expenditure has been on the rise since on donor assistance with about 40 percent of 2004. The availability of higher overall fiscal its fiscal revenue coming from donors in 2013, revenues described in the previous section has compared to about one percent on average for allowed Haiti to expand public spending and Small Caribbean States or in LAC (Table 2.2). catch up with comparator countries (Figures Nicaragua and Honduras depend on donors for 2.5 and 2.6). In 2004, Haiti’s public expenditure about 10 percent and 8 percent of their fiscal amounted to 11.4 percent of GDP, the lowest revenue, respectively. level in the LAC region. Since then, public spending increased gradually up to 2009, when 12 BETTER SPENDING, BETTER SERVICES it reached about 22 percent of GDP. This trend rise in external aid since 2004, and especially has accelerated after the earthquake, when following the earthquake. public expenditure increased to 28 percent of GDP in 2012. Even though, public expenditure D. … Despite Widening Deficits, has declined in the last two years, at 26 percent of GDP in 2014, it remains high by historical Debt is Expected to Remain standards. As shown in Figure 2.5, the main Sustainable … factor in this change has, however, been the 2.6. Haiti’s twin fiscal and current account deficits have widened. The fiscal deficit averaged 1.3 percent of GDP during the period 2005 to 2008, Figure 2.5 : Public Expenditures by Source of but widened to 7.1 percent and 6.4 percent of Financing, 2001-13 GDP in 2013 and 2014, respectively (Figure (In Percentage of GDP) 2.7). This widening fiscal deficit mainly reflected 30 transfers to the public electricity company 25 (EDH), which amounted to about 1.5 percent 20 of GDP in 2013, and the fuel retail price freeze 15 that had a fiscal cost of almost 2 percent point 10 of GDP (Box 2.1). The counterpart of the rising fiscal deficit was a rising external current 5 account deficit (Figure 2.8). Both deficits are 0 largely financed by concessional (Petrocaribe) 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 flows from Venezuela. Current - Treasury Capital - Treasury Capital - Petrocaribe Capital - Grants Sources : Ministry of Finance (MEF) and IMF Figure 2.6 : General Government Expenditure - International Comparison 2004-13 (In Percentage of GDP) 31.3 35 31.5 29.7 29.0 29.1 28.1 28.0 28.2 28.0 27.2 26.0 30 25.3 23.6 22.5 25 19.0 18.4 18.3 17.1 17.0 20 14.3 13.4 11.4 15 10 5 0 Jamaica Small Caribbean States Honduras Panama LAC Nicaragua El Salvador Dominican Rep. Costa Rica Guatemala Haiti 2004 2013 Source : IMF 13 Figure 2.7 : Central Government Fiscal Balance, Figure 2.8 : Current Account Balance, 2004-14 2004-14 (Percentage of GDP) (Percentage of GDP) 4 1 2 0 -1 0 -2 -2 -3 -4 -4 -5 -6 -6 -8 -7 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Sources : Ministry of Finance (MEF), IMF and World Bank Sources : IMF and Central Bank of HAiti (BHR) sta calculations Figure 2.9 : Haiti – Stock of Debt from Petrocaribe, Figure 2.10: External Debt (PV), 2014-35 2009-14 (Percentage of Exports G&S + Remittances) (In Percentage of GDP) 20 90 18 80 16 70 14 60 12 50 10 8 40 6 30 4 20 2 10 0 0 2008 2009 2010 2011 2012 2013 2014 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Source : IMF Sources : IMF and World Bank 2.7. While increasing, Haiti’s debt is expected to funds has allowed the financing of higher levels remain at a sustainable level. As previously of public investments and some social programs. mentioned, Haiti benefitted from significant Despite a steady accumulation of debt towards debt relief during 2009-11, with external debt Venezuela (amounting at end-2014 to about 18 declining from 30 percent of GDP in 2008 to percent of GDP and accounting for 88 percent 9 percent in 2011. The country has taken the of Haiti’s outstanding external public debt), opportunity of this greater borrowing space the latest joint Bank-Fund Debt Sustainability to contract loans on concessional terms with Analysis concludes that the country’s risk of debt Venezuela through the Petrocaribe agreement distress is moderate (Figure 2.10). (Box 2.1 and Figure 2.9). The availability of these 14 BETTER SPENDING, BETTER SERVICES Box 2.1 : Electricité d’Haiti Due to poor management, the performance of the national, vertically integrated electricity utility Electricité d’Haïti (EDH) has deteriorated over time. Worsening commercial performance has led to a lack of infrastructure maintenance and the quality of electricity service has rapidly deteriorated, including frequent service interruptions and large voltage fluctuations. The sector’s institutional framework is obsolete (e.g. not allowing EDH to tap into important renewable energy potential), sector policies are out of date, planning and monitoring of sector activities is inadequate, and vested interests have hampered reform efforts. The compound effect of the earthquake on power generation, transmission, and distribution further aggravated EDH’s weak performance by delaying key modernization activities (e.g. new billing system and the rehabilitation of the Peligre hydropower plant). In 2015, EDH’s losses stood at 57 percent of the electricity generated (of which commercial losses represented nearly 35 percent). Only one- fourth of Haiti’s population has access to electricity, and in rural areas, access is around 5 percent. EDH’s weak grid infrastructure, poor commercial performance, and inadequate controls over sub-contracted electricity generation by IPPs have led to a financial drain on Government resources. Due to its inability to meet electricity demand and in an attempt to expand electricity availability, EDH has subcontracted part of the production of electricity to IPPs.13 Unable to cover its operating expenses, including fuel costs and power purchases in part because of low bill collection rates, EDH has relied on fiscal transfers from the Treasury averaging USD200 million annually in recent years (equivalent to 10 percent of the national budget and 1-2 percent of GDP).14 Past governance reforms to preserve the integrity of public resource management in the electricity sector have produced modest results. Achievements to-date include: (i) EDH Commercial Recovery action plans (2012-13; 2013-14; and 2014-15); (ii) a National Directions Paper (endorsed by Cabinet on September 15, 2013); and (iii) a Government Energy White Paper (in draft). Since 2014, an active dialogue on the management and reform of the energy sector has been initiated within the Government’s Energy Commission and with the strategic partners in the sector. 13 IPPs provided 60 percent of total electricity generation in 2011. 14 Electricity transfers are unbounded due to their dependence on international oil prices, electricity supply, unsanctioned theft and bill non-payment, inter alia. 15 Box 2.2 : Petrocaribe and Haiti The Petrocaribe initiative, created in 2005, has 18 members.315 The agreement aims to promote cooperation among state energy operators, in terms of technology, energy policy, joint exploration, refining, sales and investment in the energy sector. The most relevant aspect of the Petrocaribe agreement is the stable supply of oil from Venezuela to other members, at favorable financing conditions. Through the agreement, Venezuela commits to providing oil to the members based on quotas established bilaterally at international market prices. Haiti joined the initiative in March 2007 and started benefiting from the agreement in October 2007, with a quota of 14 thousand barrels per day (b/d). Petrocaribe Financing Conditions Price of Barrel (dollars) Share of Financing Repayment Period (years) Grace Period (years) (% of total) 150 70 23 2 100 60 23 2 80 50 23 2 50 40 23 2 40 30 23 2 30 25 15 2 Source : BMPAD The sale of these oil products in the domestic market creates considerable resources for the government. Since 2008, the Petrocaribe flows have amounted cumulatively to 25 percent of GDP. An autonomous agency of the Ministry of Finance, (the Bureau de Monétisation du Programme d’Aide au Développement, BMPAD), manages these resources on behalf of the Haitian government. BMPAD plays an intermediary role between Haitian fuel purchasers and the Venezuelan supplier. Petrocaribe resources have been used to finance investment and social projects, as well as support the electricity sector. Whenever the government decides that new projects will be financed using these resources, it publishes a resolution listing these projects and the amounts to be financed in the official gazette. Disbursements into the projects are reported in BMPAD’s webpage on a regular basis. These projects suffer nevertheless from the same shortcomings as the rest of the public investment program, notably a lack of proper assessment, prioritization and monitoring. 15 Antigua and Barbuda, Bahamas, Belize, Cuba, Dominica, Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Nicaragua, Dominican Republic, St. Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname and Venezuela. 16 BETTER SPENDING, BETTER SERVICES Figure 2.11: International Aid, 2008-25 Figure 2.12: Government Deposits, 2009-14 (In Percentage of GDP) (In Percentage of GDP) 20 8 18 7 16 6 14 5 12 4 10 3 8 2 6 1 4 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2010 2011 2012 2013 2014 Source : IMF Sources : IMF and World Bank sta calculations E. … But Financing Constraints Are oil prices, Haiti has experienced a substantial decline in the availability of concessional Becoming Tighter financing (Figure 2.13). International oil prices 2.8. The exceptional level of aid following the 2010 are expected to remain low in the near future, earthquake is rapidly declining, however, making it even more challenging for the country eroding fiscal buffers. Donor assistance has to balance the need to ensure macroeconomic declined from its peak after the earthquake stability while protecting social and investment to 7 percent of GDP in 2014 (Figure 2.11). spending needed for social cohesion and growth. Additional fiscal revenue was not mobilized or public spending cut to offset fully this substantial F. … While Natural Disasters Would decline in resources and government deposits Call for Larger Fiscal Buffers needed to be drawn upon (Figure 2.12). In light of Haiti’s reliance on donor assistance 2.10. Natural disasters involve an extremely high to maintain the country’s fiscal position on a opportunity cost. Work was carried out jointly sustainable path, greater efforts are called for to with the Ministry of Economy and Finance mobilize more own fiscal revenue, improve the (MEF) to develop several hypothesis regarding efficiency of existing public spending programs, contingent liabilities for the government in the and increase their growth dividends, as will be event of disasters. The results suggest that the discussed in more detail in the next Chapters. annual average fiscal cost for the government 2.9. The recent decline in international oil prices is can be estimated at around one percent of GDP. further tightening fiscal resources. According This estimated cost was based on modeling used to the Petrocaribe agreement, lower international by the CCRIF and is linked to damages caused by oil prices imply a higher payment rate and less tropical hurricanes (0.8 percent of GDP) and by concessional financing. With lower international earthquakes (0.2 percent of GDP). In the absence of a financial strategy to manage this risk and 17 Figure 2.13: Petrocaribe Financing, 2008-17 (Millions of US Dollars) 30 25 20 15 10 5 0 2008 2009 2010 2011 2012 2013 2014 2015p 2016p 2017p crude at US$ 50 crude below US$ 40 crude at US$ 100 World Bank projections actual Sources : Gouvernment of Haiti and World Bank sta appropriate fiscal buffers, natural disasters will require reallocating public funds and involve high opportunity costs. 2.11. Less frequent, higher-impact events can pose a real threat to the sustainability of public finances. While frequent, low-impact events constrain fiscal space, less frequent, higher- impact events can imply significant fiscal shocks and pose a real threat to the sustainability of public finances. While the government can mobilize a portion of its financing with respect to low-impact events through budget reallocations or deposits, these funds will not be sufficient to address major events that give rise to costs that far exceed the government’s financing capacity. The fiscal shock associated with a hurricane that occurs once every 50 years, for instance, is estimated at about 10 percent of GDP, representing a much greater risk to fiscal sustainability and macroeconomic stability. 18 BETTER SPENDING, BETTER SERVICES Chapitre 3 : Protecting Priority Spending16 This chapter reviews recent trends in public spending, both by economic classification and by ministries. Haiti needs to build its human and physical capital to achieve its aspirations of becoming an emerging economy. Despite serious data limitations, the overall picture would suggest that the availability of both donor assistance and concessional financing over the past decade has allowed the country to increase spending in these priority areas. While capital spending has accelerated, operating expenses have, however, not followed, questioning the viability of many public investments. Moreover, although the wage bill has been driving the increase in current spending, most of it is accounted for by staff not directly involved in service delivery. The heavy reliance on donor and concessional financing makes social sectors particularly vulnerable to a decline in these resources. Furthermore, despite the recent increase is social spending, the role of the State in basic services such as health and education remains limited and access is inequitable. A. Haiti Has a Vision Calling for Higher and Development of Haiti (PARDH) prepared with its international development partners Human and Physical Capital …16 following the 2010 earthquake. The PSDH 3.1. Building on its opportunities, Haiti has a details the PARDH’s vision and strategic plan vision to become an emerging economy by for the country’s long-term development, and 2030. The Government of Haiti issued in May the four major areas of work for the recovery 2012 a Strategic Development Plan (PSDH), and development of Haiti (territorial reform, building on the Action Plan for the Recovery economic reform, social reform and institutional reform). The Plan aims at building a new modern, diversified, resilient, competitive and 16 Prepared by Kassia Antoine Bela da Silva, Ibrahim El inclusive economy, respectful of its environment ghandour, Eduardo Estrada, Konstantin Wacker, and and in which people’s basic needs are met. Emilio Sacerdoti. 19 Figure 3.1 : Extreme Poverty Simulations Poverty Rate Projections, 2013-2030 Poverty Rate Projections, 2013-2030 (Percentage of Population) (Percentage of Population) 23.8 23.8 24 24 20 18.0 20 16 13.8 16 13.8 12 12 13.0 8 8 3.0 3.0 4 4 0 0 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 GDP growth 3.3% (average 2011/14) GDP growth 3.3% (average 2011/14) GDP growth 2.5% (Best Spell 2005/09) GDP growth 3.6% (bottom 40% grows double GDP growth 7.2% (goal of 3% Ext. Pov.) the average) GDP growth best spell - Bottom 40% grows double Source : World Bank sta calculations Source : World Bank sta calculations This objective would require ambitious double reconstruction years (2011-2014). Assuming digit growth rates, a significant break from the unchanged income distribution, per capita GDP past, based on an expansion of agriculture, would need to grow by about 7 percent per construction, manufacturing, and tourism. This year for extreme poverty to fall to 3 percent by strategy remains, however, overly vague and 2030. This would require a two- to three-fold provides little guidance on the appropriateness acceleration in Haiti’s growth rate with respect of the composition of public spending. to its best performing years: a very ambitious outcome. 3.2. Faster economic growth alone will, however, not be enough to bring significant improvements 3.3. Policies to ensure more inclusiveness are in the living standards of most Haitians. needed. Increasing growth of the bottom 40 Simulations show that if growth in Haiti up to percent by around one percentage point more 2030 were to follow its historical performance, than the best observed average growth rate poverty would hardly decline. Under the more (2005-2009) would add considerable impetus sustained performance of one percent real per to poverty reduction. In this scenario, per capita capita growth rate observed in Haiti over the real GDP of the bottom 40 percent would grow 2005-2009, poverty reduction would still fall twice as fast as the mean. Such a performance significantly short of reaching the goal of extreme would cause poverty to decrease by 5 percentage poverty of 3 percent or less by 2030 (Figure 3.1). points more than in the distribution neutral Reducing extreme poverty by almost half in 15 growth scenario, coming closer but still falling years, to about 14 percent, would require an short from reaching the 3 percent target for overall GDP growth of 3.3 percent per year, the extreme poverty by 2030. To achieve the 3 average growth registered in post-earthquake/ percent target, a combination of faster and more 20 BETTER SPENDING, BETTER SERVICES Figure 3.2 : Economic Magnitude of Estimated Parameters (In Percentage Points) Cabinet Changes In ation In ation variation School enrollment Mobile phone use Transparency Life expectancy -0.03 0.00 0.03 0.06 0.09 0.12 bottom 20 bottom 40 mean Source : Services de la Banque mondiale inclusive growth would be needed: a growth of health (proxied by life expectancy), as well as about 4 percent per year with the income of the education (measured by school enrollment) are bottom 40 growing at twice that speed. But what important for increasing income generation at would be the measures needed to deliver this the bottom of the income distribution (Figure faster growth in lower income groups? 3.2). Reasonable changes in political stability (measured by the number of cabinet changes 3.4. Human capital and political stability matter and within-regime instability) matter slightly more for lower income groups. To provide more to the lower income groups. Infrastructure a more specific framework to assess the (measured by mobile phone usage, a common appropriateness of the composition of public measure in the literature) and transparency, spending, we built on Araujo et al. (2014) and although not disproportionally favoring the examined more closely the factors that could lowest income group, are nevertheless important lead to faster growth for lower income groups. for overall income growth. First, the correlation between potential drivers of income and the income growth of the poorest 3.5. As an alternative, Haiti’s performance in key 20 and poorest 40 percent of households were variables was benchmarked to a reference estimated, using household data for a sample of group. As a reference group, the average of an 100 countries (Box 3.1). This measure gives us aspirational group of countries was taken (Belize, the expected impact of changing one of these Dominican Republic, Ecuador, El Salvador, drivers on the income growth of the poorest. Jamaica, Paraguay, and Peru). In terms of income Second, following Araujo et al. (2014), the levels, these LAC countries broadly stand today standard deviation of each variable is taken to where Haiti aspires to be after 2030, i.e. beyond illustrate the magnitude of possible or realistic a threshold of USD 2,310. Figure 3.3 illustrates change one could expect in this variable. Haiti’s position relative to its aspirational peers Multiplying the two gives a rough estimate of the with respect to life expectancy. It shows that the expected effect of a realistic change in a variable country is in the lowest third percentile of the for income developments. The analysis finds that health distribution and lags behind the group 21 average by approximately ten years. Figure 3.4 political instability as well (especially Peru, does the same for cabinet changes, highlighting Ecuador, and Paraguay). that other peers suffered from comparable Box 3.1 : Benchmarking the Drivers of Shared Prosperity: An Application to Haiti17 We estimate economic magnitudes for key correlates of income growth at the mean, the bottom 40 percent and the bottom 20 percent of the income distribution across a set of about 100 countries. While econometric exercises investigating determinants of GDP growth have been numerous, less is known about factors influencing household incomes at different segments of the income distribution. Empirical work by Dollar and Kraay (2002) and Dollar et al. (2013) broadly rejects the idea that other factors than mean income growth would influence incomes of the poor, thus suggesting that growth would be mostly distribution neutral. Other studies have attempted, by contrast, to outline key drivers of pro-poor growth (e.g. Bourguignon, 2003; Ravallion and Chen, 2007; Christiaensen et al. 2013). Our results add to this discussion and suggest that some factors indeed allow growth to be more beneficial for the bottom 40 percent or the bottom 20 percent of the income distribution. Our data set is mainly based on Dollar et al. (2013). The data covers household data from the World Bank’s POVCALNET and the Luxembourg Income Study (LIS) databases. Household income data is organized in “spells,” i.e. income changes between two survey years, calculated as average annual log differences.18 These are calculated for average income, income at the bottom 20 percent or the bottom 40 percent. We focus on those non-overlapping spells that are at least five years long, providing 299 spells for 117 countries with a median spell length of 6 years, which is also the preferred sample of Dollar et al. (2013). To this data, we add macroeconomic variables, especially measures of political stability or institutions that are of particular interest for Haiti. Based on a fixed effect estimation, the results confirm that the income at the bottom 20 percent and bottom 40 percent grow with mean income. The estimated effect is a little smaller than previously thought, although the estimated elasticity around 0.85 is statistically not significantly different from one. We also find that better health and education outcomes, as well as lower political instability, have an economically relevant effect for income growth that goes beyond the effect via mean income (i.e. it is stronger at the bottom of the distribution. 17 Based on a background paper prepared for the SCD by Kassia Antoine, Raju Singh, and Konstantin M. Wacker. 18 POVCALNET data is either income or consumption, LIS data is disposable income. We still refer to “income” in our paper. 22 BETTER SPENDING, BETTER SERVICES Figure 3.3 : Life Expectancy at Birth, 2010 Figure 3.4 : Cabinet Changes, 2003 (Years) (10 Years Average) 80 80 1.2 120 70 70 1.0 100 60 60 0.8 80 50 50 40 40 0.6 60 30 30 0.4 40 20 20 0.2 20 10 10 0 0 0 0 Haiti Haiti El Salvador Paraguay Dominican Rep. Jamaica Belize Peru Ecauador Group LAC Belize El Salvador Jamaica Dominican Rep. Paraguay Ecuador Peru Group LAC 2010 value p-ranks (rhs) 2003 value p-ranks (rhs) Source : World Bank Source : World Bank 3.6. Again, closing the gaps in human capital and Figure 3.5 : Income E ects of Closing the Gap achieving greater political stability seem to (In Percentage of the Gap) have the highest potential for lower income groups. Multiplying the gap in the indicators Life expectancy at birth between Haiti and the reference aspirational group by the unconditional effects estimated Mobile phone use above captures the potential income gains Haiti Cabinet Changes may experience from closing the gap in these key variables. This idea, based on Araujo et Transparency al. (2014), not only highlight the areas where School enrollment the country lags behind most but implicitly In ation variation weights this gap by the economic relevance of closing it.19 Figure 3.5 shows the economic In ation income effect from closing the gap relative to the reference aspirational group for different parts 0 5 10 15 of the income distribution. Variables relating to Mean bottom 40% bottom 20% health, education, political stability clearly stand Source : World Bank out. Again, closing Haiti’s gap in infrastructure, 19 While it is implicitly assumed that a one-unit improvement in closing a large gap demands less effort than an equivalent one-unit improvement in a small gap, this might not necessarily be the case. Our exercise hence does not explicitly take into account the costs or efforts of closing this gap. 23 although not disproportionally favoring the been declining from an average of below 45 lowest income groups, is important for overall percent in 2000–03 to an estimated average of growth. Given Haiti’s recent performance in about 35 percent in 2004–06. The decline in the maintaining macroeconomic stability, the wage bill had been particularly pronounced after gap with the aspirational reference group in 2000, reflecting the freeze on nominal wages in terms of low and stable inflation is small and the public sector. At about 3 percent of GDP, the any incremental narrowing would have only a government wage bill during this period was limited effect on income growth. This does not below the levels of 6–7 percent of GDP observed mean, however, that preserving macroeconomic in other low-income countries. Looking at the stability is not important. other current spending items, the cancellation of part of Haiti`s external debt through HIPC and B. … Reflected in the Rise in Public MDRI in 2008 and further debt relief after the earthquake lowered interest payments. Outlays Investment … on transfers and subsidies, as well as government consumption of goods and services, were 3.7. The strong expansion in capital spending has broadly declining before the earthquake, but been the most striking development over the have been on the rise since. Both spending items past decade. As mentioned in the previous averaged about 2.5 percent of GDP over the Chapter, greater donor assistance and the period 2005 and 2010. This average increased to availability of concessional financing resources above 3 percent of GDP in the period following have allowed an expansion in capital spending the earthquake (2011-14). (Table 3.1). This increase also highlights a substantial shift in the country’s priorities 3.9. However, the high share of the wage bill in as reflected by the growing share of public current spending limits the role of the budget investment: it represented in 2014 more than as an effective policy tool. Even if the wage bill in half of total public spending (compared to a third terms of GDP in Haiti is in line with economies in 2005). Donor assistance and concessional at similar levels of development, it takes up financing represent, nevertheless, about 70 a significant portion of current spending, percent of the financing for public capital reflecting the country’s very low overall revenue spending, and makes this item particularly and expenditure levels. Already back in 2005, vulnerable to a decline in the availability of these total spending on wages and salaries was about resources. 35 percent of public current outlays. In 2014, this ratio had increased to almost half of Haiti’s 3.8. Wages have been driving the rise in current current budget (46 percent). The preponderance expenditure. As discussed previously, current of wages and salaries in current expenditures expenditure has been on the rise, increasing translates into more rigidity, hampering the from just below 10 percent of GDP in 2005 ability of fiscal policy to shift spending from one to about 13 percent GDP in 2014, in line with category to another if needed, as the expansion higher mobilization of own fiscal revenue. A of the payroll increases mandatory spending higher wage bill has been mostly driving this as well as future expenditure on civil servants’ development (Figure 3.6). Wages in relation to pensions. GDP rose significantly from 2005 to 2014 (from 3.5 to about 6 percent of GDP). This expansion 3.10. Furthermore, the composition of public reflects partly a catching up. The share of wages spending could also put into question the and salaries in total current expenditure had viability of many investments. As public 24 BETTER SPENDING, BETTER SERVICES Table 3.1 : Fiscal Indicators (In Percent of GDP) 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Revenues 13.1 13.5 15.8 15.1 17.8 23.9 21.9 23.4 20.8 19.7 Fiscal revenues 9.7 10.2 10.5 10.7 11.2 11.8 12.8 12.8 12.7 12.6 Domestic revenues 6.3 6.5 7.2 7.2 7.4 7.3 8.1 8.6 8.0 8.7 Customs 2.5 3.1 3.1 3.2 3.3 4.3 4.5 4.2 3.9 3.4 Other current revenues 0.8 0.6 0.3 0.4 0.4 0.2 0.3 0.1 0.8 0.5 Grants 3.5 3.3 5.2 4.4 6.7 12.1 9.1 10.6 8.1 7.0 Expenditures 13.8 15.2 15.6 18.2 22.4 21.7 25.5 28.2 28.0 26.1 Current 9.6 9.8 8.6 10.7 11.6 11.2 11.6 11.9 11.8 12.6 Wages and salaries 3.4 3.3 3.7 4.7 5.0 5.5 4.9 5.1 5.5 5.8 Goods and services 2.6 3.1 1.4 3.4 2.9 2.6 2.5 3.5 3.1 3.4 Interest payments 0.7 0.5 1.1 0.7 0.8 0.6 0.4 0.4 0.5 0.5 Internal 0.4 0.1 0.3 0.4 0.4 0.2 0.1 0.1 0.1 0.2 External 0.3 0.4 0.8 0.3 0.4 0.4 0.4 0.3 0.4 0.3 Subsides and transfers 3.0 2.8 2.4 2.0 2.9 2.5 3.8 2.9 2.8 2.9 Capital 4.1 5.4 7.0 7.5 10.8 10.5 13.9 16.3 16.1 13.5 Domestically financed 1.3 1.0 1.6 2.2 4.1 5.5 5.8 5.9 9.0 7.5 o/w Treasury 1.3 1.0 1.6 2.2 1.5 4.4 3.4 3.0 4.6 2.7 o/w Petrocaribe 0.0 0.0 0.0 0.0 2.6 1.1 2.5 2.8 4.0 3.8 Foreign-financed 2.9 4.4 5.4 5.2 6.7 5.0 8.1 10.4 7.1 6.0 Overall balance -0.6 -1.7 0.2 -3.1 -4.6 2.2 -3.6 -4.8 -7.1 -6.4 Balance exc. grants -4.1 -5.0 -5.1 -7.5 -11.3 -10.0 -12.7 -15.4 -15.2 -13.5 Adjustment (unsettled payment 0.0 0.0 0.0 0.0 -0.2 1.4 0.0 0.0 0.9 -0.1 obligation) Financing 0.6 1.7 -0.2 3.1 4.4 -0.8 3.6 4.8 8.1 6.3 External financing 0.6 2.0 0.0 2.7 3.0 3.6 5.4 4.7 4.6 4.0 Loans 1.6 1.9 0.7 2.7 3.0 3.6 5.4 4.7 4.6 4.0 Disbursments 2.4 1.9 1.5 3.4 3.7 3.7 5.4 4.7 4.8 4.3 Amortizations -0.8 0.0 -0.8 -0.7 -0.7 -0.1 0.0 0.0 -0.2 -0.3 Arrears -1.0 0.2 -0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Internal financing 0.0 -0.4 -0.6 0.0 0.8 -4.4 -1.8 0.1 3.5 2.3 Banking system 0.0 -0.3 -0.6 -0.1 0.2 -5.2 2.4 -1.1 3.8 2.7 BRH 0.0 -0.2 -0.4 0.0 0.2 -4.2 -1.0 -0.1 0.6 1.4 Commercial banks 0.0 -0.1 -0.1 -0.1 0.0 -1.0 -1.4 -0.9 3.2 1.3 Nonbank financing 0.0 -0.1 0.0 0.1 0.6 0.9 0.6 1.2 -0.3 -0.4 Arrears 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Debt reschedule 0.0 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 HIPC 0.0 0.0 0.4 0.3 0.5 0.0 0.0 0.0 0.0 0.0 Sources : Ministry of Fiannce (MEF), IMF and BOOST database *Note : Annual date covers Haiti’s fiscal yeasr (Oct. 1 to Sep. 30) 25 Figure 3.6 : Current Expenditure, 2005-14 Figure 3.7 : Operating and Capital Expenses, 2005-14 (In Percentage of GDP) (In Percentage of GDP) 7 18 6 16 14 5 12 4 10 3 8 6 2 4 1 2 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Wages and salaries Good and services Operating expenses Interest payments Subsides and transfers Capital expenses Sources : Ministry of Finance (MEF) et IMF Sources : Ministry of Finance (MEF) et IMF investment has increased, current spending has classifications, periods, and stages of spending not followed. Moreover, within this constant reported makes it extremely difficult to paint an envelope, operating costs more specifically (e.g. overall picture of spending priorities over the equipment, utilities) have remained broadly past decade and any related shifts. Nevertheless, unchanged over the past decade (Figure 3.7). patching together the disparate available data Part of the growing wage bill reflects increases for social spending suggests that the functional in support, administrative or security staff, and composition of public spending has grown closer not in staff directly delivering social services (see to the country’s priorities as described in the Chapters 6 and 7 for a more detailed discussion). first section of this Chapter. Overall, the share Without the needed resources to operate the of resources allocated to social sectors seems newly built investments and equipment, it is to have expanded from an average of about 16 questionable whether these new facilities could percent of total public spending for the period be maintained or deliver the expected growth 2007 to 2010 to about 28 percent of total public dividends. spending after the earthquake, for the period 2010 to 2012. In addition, 22 percent of donor C. … And Greater Priority Given to assistance on average is estimated to have gone towards social sectors over the period of 2010 to Social Sectors … 2012 where data are available. 3.11. Evidence suggests that greater priority 3.12. Treasury resources have been geared more has been given to social sectors. Assessing towards social sectors. Looking at each source developments in public spending according of financing in turn, one observes that the share to a functional classification is particularly of the country’s own fiscal resources managed challenging in Haiti (Box 3.2). Fragmentation by the Treasury have been over the past decade of the fiscal data in terms of different budget increasingly allocated to the social sector 26 BETTER SPENDING, BETTER SERVICES Box 3.2 : Fiscal Data Challenges Haiti does not have a consolidated financial statement. Fiscal data is widely scattered and has been a challenge for any work on public expenditure in Haiti over the recent past. As in many countries relying on aid, there is no consolidated financial statement combining current and capital spending. Information on current expenditure data is easily accessible and was extracted from Haiti’s expenditure management system, SYSDEP (Système d’Informatisation des Dépenses). The data include all the administrative units of the central government. There is no information management system, however, recording public investment expenditures in Haiti. Data for domestically-funded investment expenditure was drawn from Budget Review Laws (Lois de Règlement) and execution reports for the Public Investment Program (PIP) (Bilan d’Execution du PIP) provided by the Ministry of Economy and Finance and the Ministry of Planning and External Cooperation. The data includes projects financed from domestic sources for all the administrative units of the central government. For foreign-financed investments, though with some limitations, sectoral data exist on commitment and disbursements, collected by the Ministry of Planning and External Cooperation through the External Assistance Management Module (Module de Gestion de l’Aide Externe - MGAE) set up in 2009. Budget classifications are problematic. Haiti’s budget classification system does not contain an explicit module for the functional classification of expenditures. This means that the information system is unable to produce reports linking each budget line with a given function. Like in many countries the administration classification has to be used as a proxy. Moreover, a large portion of externally-financed expenditures is executed outside the budget system, with donors using their own implementation arrangements. As a result, the investment budget follows different classification depending on the sources of financing (Treasury i.e. own resources, Petrocaribe, and donors). Furthermore, on the domestically-funded part of the investment budget, there is no information on actual payments. While current spending can be followed from budget allocation to payment, only information up to the disbursements for specific domestically-funded projects is available. Recorded disbursements (décaissement) are transfers made to bank accounts held by individual line ministries for each project and hence do not necessarily imply final project expenditure execution. The information on the overall position of the government in the banking system tells us what level of spending has actually taken place and allows a discussion on broad trends according to the economic classification. Discussing trends along functional lines within this overall spending is much more challenging, however, because information on individual project account balances is not yet available. The ongoing rolling out the Single Treasury Account is expected to provide more detailed information on the payments made out of the project accounts and allow for a more detailed reporting of investment expenditure in the future. 27 Figure 3.8 : Changes in Sectoral Composition – Figure 3.9 : Project Activities Financed by Petrocaribe (In Percent of Total Treasury Spending) Funds, 2008-13, (USD Millions) 16 360 12 300 8 240 4 0 180 -4 120 -8 60 -12 -16 0 2007-2009 2009-2011 2011-2013 2007-2013 2008 2009 2010 2011 2012 2013 Infrastructure Agriculture Others Economic Sector Political Sector Territories Water & Sanitation Economy Social Sector Culture Sector Energy Housing Social Sources : Ministry of Finance (MEF) et IMF Source : BMPAD (including the ministries of health, education by these resources (Figures 3.9 and 3.10). From and social affairs). This average hides, however, 2008 to 2013, Petrocaribe funds have been mostly wide differences (Figure 3.8). The period before allocated to infrastructure projects (58 percent), the earthquake (2007-2009) seems to have followed by social programs (9 percent), housing been characterized by the need to strengthen (7 percent) and agriculture (6 percent). The institutions and human capital with the social availability of Petrocaribe resources channeled to and political sectors seeing an increase in infrastructure may explain partly why Treasury their share of spending (the latter includes the financing shifted away from this sector. ministries of justice and interior). The earthquake 3.14. Donor assistance seems also to have broadly triggered a clear shift in priorities towards favored infrastructure. Neither the Treasury the economic sector (including ministries in nor the MEF monitor the expenditures financed charge of reconstruction such as the ministries by donors. The Ministry of Planning and of planning and public works). More recently External Cooperation aggregates information (2011-2013), with the need of reconstruction on the projects financed by donors through the receding, the social and political sectors have Module de Gestion de l’Aide Externe (MGAE) regained attention. set up in 2009 and started recording data from 3.13. Petrocaribe funds are reported to have been 2010 onwards. The MGAE system has its own directed to infrastructure and, to a lesser extent, limitations, however, as MGAE data does not social programs. Data on Petrocaribe made always match data on official transfers in the available by BMPAD, even though it does not Balance of Payment, or to the data reported by follow administrative or functional classification, the IMF. Nevertheless, the data reported through gives an idea of the investment activities financed the MGAE helps to reach a more detailed view 28 BETTER SPENDING, BETTER SERVICES Figure 3.10: Project Activities Financed by Petrocaribe Figure 3.11: Donors Financing, 2010-12 Fund, 2008-13, (Percentage of Total) (Percentage of Total) Social Transport Housing Energie Health 9% 12% 7% 4% 6% Agriculture Others Policies on health 6% 1% and population Others 2% Debt related 7% Non speci ed 19% Territories 8% In-kind Infrastruct. 2% Infra. & Social 9% 58% Services Water & 5% Agriculture Sanitation Gov. & Civil Society 4% 4% 7% Financial Economy Entreprises services 2% 1% 2% Energy 5% Education Water & Multi-sector 9% Sanitation 6% 7% Source : BMPAD Sources : Foreing Aid Management Database (MGAE) and World Bank Sta calculations of the composition of public expenditure. ratio has expanded to 45 percent in 2012. This From 2010 to 2012, external aid has covered a heavy reliance on donor financing makes social broad range of sectors, but seems to have been sectors particularly vulnerable to the decline in mainly channeled to activities in infrastructure donor assistance and could put into question the (infrastructure and transport), health and sustainability of some programs (see Chapters 6, education (Figure 3.11). 20 7 and 8 for a more detailed discussion). 3.15. The reliance of social sectors on donor assistance for their financing seems to have D. … But Haiti’s Inclusiveness increased. While the availability of donor Remains Limited assistance has allowed for financing of an expansion in social sectors, it has also increased 3.16. Despite the recent increase in social public the reliance of these sectors on donor financing. spending, Haiti’s economic system is not very Social sectors including health, education and inclusive. The Bank’s 1998 Poverty Report social protection are considered the most aid- noted that “Haiti has never had a tradition of dependent sectors in the country. Overall, governance aimed at providing services to the social sectors relied on donor assistance for population or creating an environment conducive 16 percent of their financing in 2010.21 This to sustainable growth.” A number of Haitian and international observers broadly agree that the Haitian State remains ineffective and delivers 20 Debt relief is another important component (14 percent). little to its population. “Haitian institutions have 21 Social sectors have been defined for this purpose as including the following MGAE classifications: education, infrastructure and social services, policies on health and population, and health. 29 and state capacity to effectively formulate and Figure 3.12: International Comparison - Social implement sound policies, and to deliver core Spending, 2013 or latest (In Percentage of GDP) public services to the population, are weak. Furthermore, the state is present largely in the 25 major urban centers and has been unable to 20 provide basic services or infrastructure to large portions of the population” (Buss, 2013). 15 3.17. Social spending remains limited and the 10 delivery of basic services highly inequitable. 5 Despite the recent increases in public spending in social sectors, public spending in health, 0 education, and social protection amounts to 5 Costa Rica (2011) Honduras (2012) Panama (2011) El Salvador (2012) Nicaragua (2011) Guatemala (2011) Haiti (2013) percent of GDP, below comparator countries, limiting the government’s ability to offer equal opportunities to its citizens (Figure 3.12). At the same time, many large spending items such Sources : World Bank and Ministry of Finance(MEF) as fuel subsidies clearly favor the rich (Chapter 9). In the absence of government, basic services such as health and education are mainly provided by non-government actors (Chapters 6 and 7). never provided justice, education or healthcare Seventy to eighty percent of primary school to the majority of the population” (Lockhart and students attend non-public schools, placing Forman, 2013). Instead, a small economic elite a substantial financial burden on households has supported a state that makes only negligible and delivering achievements closely linked investments in human resources and basic with household income. Outcomes are equally infrastructure (World Bank, 1998). “Governance unfavorable to the poor in the health sector. 30 BETTER SPENDING, BETTER SERVICES Chapter 4: Towards greater and more equitable revenue mobilization22 This Chapter reviews recent trends in fiscal revenue. Haiti has managed over the past decade to increase its fiscal revenue mobilization ratio. This ratio remains, nevertheless, relatively low, revenues originate mainly from the Metropolitan area or at the borders, and the overall tax system is regressive. Exemptions are eroding the tax base, and would call for a review and close monitoring. Personal income tax thresholds do not reflect the country’s socio-economic structure with the highest marginal rate applied only to a small number of taxpayers. In addition, although Haiti has very low trade tariffs, it levies substantial inspection fees. A. Despite Some Increase, Revenue addition, a new tax instrument was introduced in 2012: the Fond National d’Education (FNE) Mobilization Remains Low and which is financed through taxes on international Regressive… 22 telephone calls and money transfers with the purpose to support the Programme de 4.1. While own fiscal revenue has increased over Scolarisation Gratuite et Universelle (see Chapter the past decade, Haiti has still one of the lowest 7). Despite this progress, Haiti has still the second revenue mobilization rates in the region. As lowest revenue-to-GDP ratio for all countries in mentioned in Chapter 2, fiscal revenue as share the region (Figure 4.2). Increasing the tax-to- of GDP both from domestic sources and trade GDP ratio is a common part of moving towards has increased over the recent past (Figure 4.1). higher levels of per capita GDP. Cross-country This improvement has been driven primarily data nevertheless suggest that Haiti currently by an increase in domestic income and sales performs slightly below average when compared tax revenue collected in Port-au-Prince. In to countries with similar per capita income levels. 22 Prepared by Jan Loeprick and Erik von Uexkull. 31 Figure 4.1 : Fiscal Revenue, 2009-13 (In Percent of GDP) 14 12 Other (inc. FNE) 10 Customs duties (inc. inspection fees) 8 Other taxes Sales tax 6 Excise taxes 4 Domestic taxes in provinces 2 Income taxes (P-au-P only) 0 2009 2010 2011 2012 2013 Sources : Ministry of Finances (MEF), IMF, BOOST database and authors’ estimate 2009 2010 2011 2012 2013 Fiscal Revenue 11.2 11.8 12.8 12.8 12.7 Domestic taxes 7.4 7.3 8.1 8.6 8.0 Income taxes (P-au-P only) 2.3 2.2 2.5 3.0 2.6 Domestic taxes in provinces 0.5 0.4 0.4 0.5 0.5 Excise taxes 0.7 0.5 0.3 0.3 0.3 Sales tax 3.5 3.2 3.7 3.7 3.7 Other taxes (P-au-P only, inc. discrepancies) 0.5 0.9 1.2 1.0 1.0 Customs duties (inc. inspection fees) 3.3 4.3 4.5 4.2 3.9 Other (inc. FNE) 0.4 0.2 0.3 0.1 0.8 Sources: Ministry of Finance (MEF), IMF and BOOST database heavily on indirect taxes which affect consumers burden of taxes weighs differently on economic independently of their income level. The ratio of agents of varying income levels. Moreover, while direct to indirect taxes stood at about 30 percent direct taxes tend to be harder to administer, they in 2011, a level inferior to that of most countries are frequently seen as an important indicator in LAC and to the average of Low-Income and tool of successful state-building as they tend Countries, and is largely explained by the fact to create a more direct citizen-state interaction that a sizable share of Haiti’s revenues come from and voice (Di John, 2010; OECD 2014). international trade (Figure 4.3). Tax systems 4.4. Compared to other countries in the region, relying relatively more on direct taxes tend to Haiti relies heavily on customs duties for be more progressive because in such systems the 32 BETTER SPENDING, BETTER SERVICES Figure 4.2 : Tax-to-GDP Ratio (a) Tax / GDP Regional Comparison (b) Tax-to-GDP relative to Per Capita GDP (Percent) (Percent and Log of PC Income) 30 35 25 30 20 15 25 Tax revenue / GDP 10 20 5 15 0 Haiti 10 Guatemala Haiti Dominican Rep. Costa Rica El Salvador Nicaragua Honduras Suriname Belize St. Lucia Dominica Barbados Jamaica Trinidad and Tobago 5 7 8 9 10 11 12 log naturel du PIB par hab. Sources : Haitian Revenue Authorities and World Bank, Sources : Haitian Revenue Authorities and World Bank, World Development Indicateurs. World Development Indicators. 4.2. Furthermore, the regional distribution of revenue generation. Revenue from customs revenue collection is heavily skewed towards duties and a 5 percent inspection fee collected Port-au-Prince and border posts. The on imports amount to about 4 percent of GDP, Tableau des Opérations Financières de l’Etat more than twice the regional average of 1.8 (TOFE) currently only provides a rudimentary percent of GDP, and also significantly above breakdown of the geographic origin of fiscal the ratios of neighboring countries with similar revenues in the sense that all revenues collected overall revenue-to-GDP ratios like Costa Rica in the provinces are lumped together in one and Dominican Republic (both at 1.1 percent category with no distinction by instrument. of GDP). Streamlining trade policy and its This limited data indicates that in 2013 only application with a view to reduce distortive approximately 4 percent of total revenue was effects, and enhance border efficiency and collected in the provinces. While the largest transparency will be important to ensure future share of revenue (60 percent) is collected at the growth and productivity, and a number of trade border through customs duties, inspection fees, policy reforms are currently being debated. and sales and excise taxes on imported goods, However, given the importance of tariffs and the remaining 36 percent are collected in Port- fees collected at the border, such reforms must au-Prince. also be guided by a careful assessment of their potential effects on fiscal revenue, a topic which 4.3. Haiti’s tax system generates not only limited is further explored in this chapter. resources for the government, but also tends to be regressive. Haiti’s fiscal revenues rely 33 Figure 4.3 : Ratio Direct To Indirect Taxation, 2009 or 2011 (Percent) Panama Low income countries Jamaica Costa Rica Honduras Dominican Rep. Nicaragua Haiti 0 20 40 60 80 Sources : IMF and World Bank B. … With Exemptions and Ill- with challenging investment climates. Tax incentives to attract investment to developing Designed Tax Brackets Eroding countries are often seen as needed to compensate Direct Taxes… investors for unfavorable investment conditions and limited provision of public goods. 4.5. A high share of corporate income tax revenue International experience suggests, however, is lost because of exemptions. Corporate and that this ‘substitution effect’ often does not personal income tax rates are largely comparable materialize in practice and that tax incentives to the regional average (Figure 4.4) and cannot can create windfalls for investments that would explain Haiti’s low tax burden. For instance, have taken place in any case (Box 4.1). According Haiti’s corporate income tax rate of 30 percent to the World Bank’s Doing Business Indicators, is slightly above the regional average (the dotted Haiti ranks particularly low on indicators such line) and above the rates in most regional as starting a business (187th out of 189) and comparators with higher tax revenue-to-GDP protecting investors (170th) that are of strong ratios. Exemptions, however, are estimated to concerns for potential investors, but can be have amounted in 2011 to the equivalent of about resolved through policy reforms that are less 63 percent of all income tax collected.23 Most costly and potentially more effective in attracting exemptions derive from the 2002 investment investors. In addition, IMF (2013) identifies code that grants 15-year-tax exemptions for several aspects of the current tax code – rate and companies in free zones as well as 5-10 year calculation method of the tax on dividends, non- exemptions for specific investment projects that deductibility of foreign taxes on international are considered desirable from a development income, strong audit requirements, and high perspective. upfront-payments of taxes – that if reformed are 4.6. International experience suggests that tax likely to have a positive impact on investment. exemptions should be used with caution as a Under these circumstances, reforms to improve means to attract FDI, especially in countries the business climate are likely to have a stronger impact in attracting FDI without the adverse of tax exemptions. 23 IMF (2013) 34 BETTER SPENDING, BETTER SERVICES Figure 4.4 : Corporate (left) and Personal (right) Income Tax Rates in Regional Comparison (Percent) 40 40 35 35 30 30 25 25 20 20 15 15 10 10 5 5 0 0 Guatemala Barbados Belize Dominican Rep. El Salvador Honduras Trinidad and Tobago Costa Rica Dominica Haiti Nicaragua Jamaica Suriname Suriname Costa Rica El Salvador Haiti Nicaragua Dominican Rep. Guatemala Honduras Dominica Barbados Belize Jamaica Trinidad and Tobago Source : USAID ‘Collecting Taxes’ 2013 dataset. Source : USAID ‘Collecting Taxes’ 2013 dataset. Box 4.1 : International Experience with Tax Incentives Every investment incentive policy has potential costs and benefits. Benefits may arise through higher revenue from possibly increased investment and social benefits such as jobs, positive externalities, and signaling effects from this increased investment. The costs are due to revenue losses from investments that would have been made even without the incentives and indirect costs such as economic distortions and administrative and leakage costs. It is difficult to quantify these elements, but trying to do so provides a useful conceptual tool for policymakers analyzing the general framework for incentives as well as targeted incentives for anchor investments, export- oriented and mobile investments, extractive industries, and so on. A key factor in the cost – benefit equation is the question of additionality (‘Would a given investment have occurred in any case had no incentive been granted?’). If the answer to this question is yes, the benefits that can be observed in relation with the investment cannot be associated with the incentive that was provided. On the other hand, the costs of the incentive are real regardless of whether the incentive was effective or not. While most international studies find a significant negative relationship between the marginal tax rate and investment inflows in developed economies, results for developing countries differ. The balance of evidence suggests that, for many developing countries, fiscal incentives do not effectively counterbalance unattractive investment conditions such as poor infrastructure, macroeconomic instability, and weak governance and markets. The Figure below shows that for countries with weak investment climates, a lower marginal effective tax rate (METR) has limited impact on FDI. For example, having an METR of 20 percent instead of 40 percent raises FDI by 1 percent of GDP for countries ranked in the bottom half in terms of investment climate—while the same difference in METR has an effect eight times greater for countries in the top half. 35 E ectiveness of Tax Incentives for Investment Attraction 30 High IC countries HN U Trend High IC countries HG K G BR Low IC Countries 20 Trend Low IC Countries FDI as % of GDP E GO L ND Almost no impact of I L S G SP D MG CE H J R O lowering E ective Tax 10 HV R I L R VM N A KZ Rates on FDI in Low Income Countries EY G TD C CL H ZB M A CN L A V L O S FI J R SB J M A CI R B GR UR K H GA R FA RM O SE L O PL CE Z U MS E PR FN I TN U Y MS SK V RS U S EP UA G TA Z H TA DK N A MR PK A G NA TR U KN E W SE BA R PT R E MX ZF A NL Z W RA I D N I A T S UA I N D DU E 0 Z UB EH T NR O R GC J N P EU C R I N KR O BA W -20 0 20 40 60 METR To confirm this finding, the World Bank’s Investment Climate Advisory conducted econometric studies and surveys of investors in developing countries. The econometric studies overwhelmingly conclude that the investment climate is more important than tax breaks or other nontax incentives. The surveys found that factors related to the investment climate—such as ease of import and export, availability of local suppliers, regulatory framework, adequate infrastructure, and the country’s geographic location—rated higher than incentives as a primary motivation for investment. For Thailand, 81 percent of investments would have been made even without incentives. In Jordan, Mozambique, and Serbia 70 percent or more of investments would have been made anyway, so incentives were redundant. Such surveys also reveal, however, that tax incentives are more important for exporters who operate in tight international competition and tend to be internationally mobile. Source: James (2013) 4.7. Where tax exemptions or other investment administrative costs, encouragement of rent incentives are used, a clear framework needs seeking behavior, and creation of economic to be in place to measure and track their costs. distortions (James, 2013). Work is in progress in Haiti to measure more 4.8. At the same time, it is important to track and thoroughly and track foregone revenue from monitor the benefits of tax exemptions. In tax exemptions. This is a welcome step. In the order for the policy maker to decide whether absence of such information, decision makers exempting an investor was a good decision may not fully factor costs from taxes that were or not, not only the amount of the foregone never collected (and thus do not appear in the fiscal revenue should be identified, but also the budget) into their decisions. In addition to expected benefits regularly monitored. This foregone revenue, tax exemptions can also have applies to benefits such as employment creation, indirect negative effects, for instance through technology transfer, or demonstration effects. In 36 BETTER SPENDING, BETTER SERVICES Box 4.2 : Removing Nuisance Taxes So called nuisance taxes are instruments that raise very little revenue, but can impose substantial compliance burdens on companies. Administering a large number of different instruments for the same taxpayer population is by definition inefficient, and often the revenue generated by such taxes barely covers their administrative cost. Anecdotal evidence suggests that a number of nuisance taxes currently exist in Haiti. In order to reduce compliance cost of firms and use limited tax administration capacity more efficiently, it is therefore important to complement tax administration reform efforts with a review of tax instruments to identify and remove such nuisance taxes. International experience suggests that despite negligible revenue collected from nuisance taxes, a reform tends to be a difficult undertaking as it can involve various entities and stakeholder with often conflicting interests. The overall reform objective should be the combination of as many taxes applied to the business sector as possible into a single tax administered and collected by the same agency. Such a reform typically requires: i. thorough review of all current instruments, their tax base, legal basis, and revenue collected. Such an exhaustive inventory is needed for all subsequent decisions on the removal and streamlining of instruments; ii. the definition of clear criteria to streamline these instruments. For example, all taxes which serve solely to generate revenue and are not directly related to a service could be assessed based on the average revenue generated by each instrument over the last 3 years. When intakes have been negligible and below a certain threshold, their elimination could be envisaged; and iii. the development of transfer mechanisms to ensure sufficient resource flows to line ministries and local institutions that previously levied their own taxes. In this process, user charges that are directly related to a service, or fulfill a regulatory function would have to be distinguished from those that only serve to generate revenue. Such taxes related to direct services or a regulatory function would have to be evaluated against more sophisticated criteria (i.e. justification of regulatory function, necessity/quality of services provided). In a second phase, taxes would actually be consolidated to reduce significantly the number of instruments and thus the interaction with tax officials. The most drastic reform option for achieving this is based on the reversal of the burden of proof for the existence of a tax instrument.24 This approach has been pioneered in Sweden and subsequently adopted in several countries (Hungary, Mexico, Kenya), and has been further developed by the World Bank’s Foreign Investment Advisory Service for reform licensing regulation. A reversal of the burden of proof would mean that line ministries, local authorities and other “owners” of tax instruments would need to prove that their taxes match pre-defined criteria for tax instruments. If they fail to justify their taxes, or do not notify its existence past a set deadline, these taxes would be eliminated by default. This mechanism reverses the incentives to share data and information required to identify nuisance taxes. All ‘surviving’ taxes would be listed in a centralized registry. The registry would have positive legal security: no regulation not in the registry can be enforced against a business. 24 The following discussion of the reversal of proof principle draws upon FIAS (2010) 37 doing so, a critical step of analysis is to determine up to HTG 250,000; 25 percent up to HTG whether a given investment is actually a result of 750,000; and 30 percent for everything above that the tax exemption, or would have occurred in (Figure 4.5). These thresholds are high relative to any case. per capita GDP compared to other countries in the region: The maximum rate is only levied at 4.9. Furthermore, anecdotal evidence suggests that incomes above more than 2000 percent of per a number of taxes exist in Haiti that generate capita GDP, more than twice the ratio for the little revenue but impose high compliance next highest comparator country (Guatemala cost on businesses (“nuisance” taxes). While at 757 percent). In other words, only very high such taxes generate very little revenue, they incomes are taxed at the maximum rate. Even can have adverse effects on the tax system as a the threshold for the second highest income tax whole by encouraging informality and adding band at 25 percent exceeds the regional average complexity to tax administration. Removing for applying the maximum rate (the dotted line), nuisance taxes can therefore be an effective and the majority of regional comparators apply complement to ongoing reform efforts to thresholds that even fall below the third highest strengthen tax administration, while at the same tax band of 15 percent in Haiti. time contributing to a more favorable business environment. Box 4.2 presents international 4.11. Household data suggests that most fiscal experience for the identification and removal of revenue could be generated at the top end nuisance taxes. of the income distribution. Figure 4.6 shows average monetary income by 0.2 percentiles of 4.10. Personal income tax rates are in line with the the working age (15+) population. 62 percent of regional average, but apply only to very few the working age population report no monetary taxpayers. There are currently four tax bands income at all. Another 14.4 percent report for personal income tax: 10 percent on incomes monetary income below the minimum income above HTG 20,000 and below 100,000; 15 percent tax threshold of HTG 20,000. 16.2 percent report Figure 4.5 : Income Tax Rate Thresholds In Regional Comparison (Per Capita GDP) 2100 1800 1500 1200 900 600 300 0 Belize 25% (max) Haiti 15% Haiti 25% Haiti 30% (max) El Salvador 30% (max) Barbados 35% (max) Haiti 10% (min) Suriname 38% (max) Jamaica 25% (max) Dominica 36% (max) Costa Rica 15% (max) Dominican Rep. 25% (max) Nicaragua 30% (max) Honduras 25% (max) Guatemala 31% (max) Sources : USAID ‘Collecting Taxes’ 2013 dataset and calculation by authors. 38 BETTER SPENDING, BETTER SERVICES Figure 4.6 : Hypothetical Income Tax Revenue Collection Based on 2012 Household Survey Results (Tax Rate in Percent and Income in Millions of Gourdes) 50 45.7 percent of 2.5 revenue from 30 % 45 income tax bracket 40 24.2 percent of revenue 2.0 from 25 % income tax 35 Millions of gourdes bracket 30 1.5 percentage 19.6 percent of revenue 25 10.5 percent of revenue from 15% income tax from 15% income tax bracket 20 bracket 1.0 15 10 0.5 5 0 0.0 75 80 85 90 95 100 Applicable income tax rate, percent (lhs) average annual labor and non-labor monetary income, G$ millions (rhs) Source : Authors’ calculation based on data from the ECVMAS 2012 monetary income that would put them in the 10 to be collected mainly on incomes in the formal percent income tax bracket, and the equivalent sectors and public service. In addition, household shares are 5.2 percent, 1.6 percent, and 0.6 survey data tends to be less accurate at the top percent respectively for the 15, 25, and 30 percent end of the income distribution where outliers brackets. If perfect income tax collection and in terms of very high incomes are not easily no strategic response in taxpayer behavior were captured. Figure 4.6 is thus merely an illustration to be assumed, this would imply that 11 percent of how the income tax brackets are aligned to of income tax revenue would be generated by the socio-economic structure of the population. taxpayers in the 10 percent bracket, and 20, 24, A more thorough review of income tax brackets and 46 percent respectively by taxpayers in the would rely on a combination of household survey 15, 25 and 30 percent brackets.25 This calculation data and actual income tax information to factor illustrates the importance of ensuring the actual in collection efficiency, and gain a more accurate collection of income tax revenue in particular at picture of top income earners. the top end of the income distribution. Perfect 4.12. There might be potential for expanding tax collection is a strong assumption that is revenue mobilization from the top 10 percent unlikely to hold in practice as income taxes tend of the income distribution by reviewing tax brackets. Figure 4.7 shows results of a simple illustrative calculation based on the 2012 25 The strategic response by businesses and individuals to tax policy thresholds has been investigated in household survey, as in the previous paragraph a number of studies. See for instance, Kleven and based on the assumption of full tax compliance Waseem (2013) and Bruhn and Loeprick (2014) and no strategic response in taxpayer behavior, looking at thresholds in Pakistan and Georgia. 39 Figure 4.7 : Hypothetical E ect of Changes in Income Tax Brackets on Revenue Collection by Income Group 120 100 80 61.5 55.5 59.2 +3.9% + 13.2% 60 total total revenue revenue 40 31.0 25.3 25.5 20 12.6 12.6 14.0 0 6.7 6.7 6.7 current system : HTG 20,000 (10%) ; example 1 : HTG 20,000 (10%) ; HTG example 2: HTG20,000 (10%); HTG100,000 (15%) ; HTG 250,000 100,000 (15%) ; HTG 250,000 (25 %) ; HTG75,000 (15%); HTG150,000 (20%); (25%) ; HTG750,000 (30%) HTG 350,000 (30%) HTG200,000 (25%); HTG350,000 (30%) 76-90th 91-95th 96-99th top 1 Source : Authors’ calculation based on 2012/13 Haiti Household Survey data. and serving for illustrative purposes only. 4.13. Further analytical work could help review Two reform scenarios are applied: Example 1 income tax brackets under equity and simply lowers the threshold for the 30 percent efficiency perspectives. While household data maximum rate from HTG 750,000 to HTG can give an indication of the alignment of tax 350,000. The effect is a relatively small increase brackets with the country’s socio-economic in overall revenue collection (about 4 percent) reality, further analysis of income tax collection which is borne entirely by tax payers in the top could focus on reviewing tax collection data 1 and 96-99th percentile of income distribution by income groups. Such analysis could aim that are moved from the 25 to the 30 percent to identify measures of collection efficiency, bracket. A more extensive reform scenario is in particular for high income taxpayers with analyzed as example 2, which includes not only high revenue collection potential, and inform a the same threshold for the 30 percent band potential review of income tax brackets in terms as example 1, but also a general lowering of of equity (ensuring a ‘fair’ distribution of the minimum thresholds with the exception of the tax burden) and efficiency (revenue generation HTG 20,000 threshold for the 10 percent rate. relative to administrative cost). It also introduces a fourth bracket at 20 percent to close the relatively large gap between the 15 C. … And Indirect Taxes Hampered by and 25 percent bands. The effect is much more substantial in terms of the additional revenue Weak Revenue Administration and generation (+ 13 percent). In terms of its Unclear Exemptions distributional impact, taxpayers below the 90th income percentile would still remain unaffected, 4.14. Customs administration seems to suffer from but the impact is spread more evenly across significant performance shortfalls. Since about taxpayers in the top 10 percentile. 60 percent of all tax revenue is collected by the 40 BETTER SPENDING, BETTER SERVICES customs authorities at the border, the ability Figure 4.8 : Results of Haiti Customs Assessment: of the customs to collect these taxes efficiently Distance from and effectively is critical for overall revenue International Best Practice mobilization. A recent World Bank analysis process using the Customs Assessment Trade Toolkit orientation (CATT) concluded that customs procedures in 100 Haiti were moving closer to international good 80 strategic practices (World Bank, 2014). Critical indicators transparency 60 thinking of practical performance such as efficiency, 40 degree of control, and trade facilitation were, 20 0 nevertheless, underperforming (Figure 4.8 facilitation control and Box 4.3). The CATT analysis highlighted several areas of recent or ongoing improvement, and an ambitious reform and modernization agenda is currently under implementation. e ectiveness e ciency Among the priorities for improvement in the Source : World Bank (2014) CATT report are the better systemization of available information, enhanced coordination between sectors, the development of a manual of operational procedures, improvement of internal shortfall rises to 50 percent (Table 4.1). The communication, and upgrading of available bulk of foregone TCA revenue is from products infrastructure. The customs administration relies that are either industrial inputs (21 percent of on the globally used SYDONIA (ASYCUDA in revenue lost) or capital equipment (23 percent of English) software for tracking and managing revenue lost). Exemptions on capital goods can customs data, but some of its more advanced be beneficial in terms of promoting investment, functions are not fully exploited, and transactions while exemptions on industrial inputs can be are not always registered in real time. absorbed as windfall profits rather than passed on to consumers if competition is weak. As with 4.15. Substantial fiscal revenue shortfalls also business tax exemptions, carefully monitoring occur at the border. While no complete data the costs versus the expected benefits of is available on TCA and excise tax exemptions, such exemptions is key for policy makers to detailed transaction level data on taxes collected understand their full impact. at the border makes it possible to calculate the value of exemptions on imported products. The 4.16. The classification of exemptions in customs foregone revenue can be estimated by comparing data is insufficient to thoroughly track their the actual revenue collected to the amount an justification. Typically, customs procedure importer should have normally paid, had the codes would give a clear indication of the legal official rate been applied. The results suggest basis for the exemption (i.e. import to free zone, that In 2013 Haiti gave up the equivalent to exemption under a specific paragraph in the 14 percent of total sales tax revenue collected investment code) that makes it possible to use at the border, plus another 19 percent in ad this data to track the costs of specific exemption valorem excise taxes.26 For tariffs, the reserve rules. In the case of Haiti, most exempted imports are simply classified as “exempt of all taxes” or 26 Exemptions on non-ad valorem excise taxes cannot be calculated with available data, but these taxes account for a relatively small share of total revenue. 41 Box 4.3 : The CATT Methodology for Benchmarking Customs Performance The Customs Assessment Trade Toolkit (CATT) is an integrated tool for measuring customs performance across countries and over time. The tool provides relevant information on customs performance by collecting data and information from different countries against a set of 120 high level indicators related to customs’ operations and producing performance benchmarks. With these, the tool can be used to evaluate the relative strengths and weaknesses of any customs administration and compare them to good practices. The distinctive feature of the CATT is that all data is objectively verifiable and thus measurements are comparable across countries and across time periods. The CATT does not simply measure the formal use of good practices, as it might be possible that most customs administrations around the world operate within a legal framework that promotes the use of simplified procedures. The CATT also considers the performance of a customs office, such as how long in fact customs administrations take to control declarations. To address this question, the real time that goods take to be released is measured. To get this information, the CATT evaluation exercises reviews to determine how the customs administration operated in practice. For example, the release of goods is compared against reasonable international standards. The CATT team has picked good practices from the existing literature on customs good practices in a wide range of areas, including simplified procedures; selective controls; risk analysis; advanced data/entry level; electronic data submission; increased use of information technology; improved management of special customs procedures; enhanced transparency and partnership with private sector; infrastructure; process orientation; and public service orientation. The 120 measurements (called high level benchmark indicators) derived from these areas are grouped into seven dimensions (Process Orientation, Strategic Thinking, Control, Efficiency, Effectiveness, Facilitation, Transparency) that summarize the overall state of a customs office, creating a quantifiable view of the overall practice and performance of a customs office. Source and further information: www.customscatt.org exempt of specific tax instruments. Exemptions D. … But Could Be Bolstered By a attributed to foreign embassies, NGO and Simple VAT government entities account for only a small share of total exemptions. It is also noteworthy 4.17. Important efficiency gains could be expected that close to 5 percent of exemptions are granted from a move towards a regular VAT instead on imports that are classified as “fully taxable”. of the turnover tax (TCA) currently applied. Improving the classification of exemptions to Because the deductibility for the TCA is limited, more thoroughly explain the justification for a the instrument can be applied to consecutive given exemption would be an important first step sales, including inputs. When the input taxes for better monitoring of exemption costs. The are not refunded, this leads to tax accumulation same principles discussed above regarding the and tax on tax (cascading). Depending on the costs and benefits of TCA and excise exemptions production chain, the cumulative tax burden apply to tariff exemptions. could thus become very high in cases where no 42 BETTER SPENDING, BETTER SERVICES deductibility is allowed, which renders affected Figure 4.9 : Tax Revenue Collection around the goods and services less-competitive. A particular Introduction of VAT concern is that the non-deductibility of domestic (Percentage of GDP and of Total Revenues) inputs sourced by exporters undermines efforts to promote linkages and the goal of strengthening 16 Introduction 36 the overall competitiveness of the Haitian of VTA 35 export sector. Technical work to determine the 15 34 33 transition of the TCA towards a regular VAT (e.g. 32 removing deductibility restrictions, building 14 31 a sound refund administration, introducing a 30 zero rate for exporters) and a schedule for its 13 29 implementation is currently under way. 28 12 27 4.18. International experience suggests that -3 -2 -1 0 1 2 3 introducing a VAT can lead to substantial Tax revenue, % of GDP (lhs) increases in tax revenue for low and lower Revenue from taxes on goods and services, % of total (rhs) middle income countries if it is designed properly. Data from a sample of nine low Source : Author’s calculation based on data from World Bank/ World Development Indictors and USAID 'Collecting Taxes' 2013 dataset and lower middle income countries that have introduced VATs since the 1990s is summarized in Figure 4.9. It shows that on average, tax revenue as a share of GDP increased from around 13 to significantly higher mandatory registration 14.6 percent of GDP within three years after the threshold to limit the number of taxpayers in the introduction of the VAT. As VATs are sometimes regime. More generally, the establishment of self- introduced in the context of larger tax reforms, it assessment procedures, effective enforcement is also important to note that the share of taxes tools, a functioning refund-process for exporters on goods and services also increased significantly and sufficient capacity to perform audits are over the same time period, indicating that VAT critical for the successful implementation of a revenue is indeed contributing to the positive VAT (Grandcolas, 2005). change. However, there is substantial variation in 4.20. Its design should be simple. General the data, suggesting that these positive outcomes recommendations in designing VAT systems are not automatic with the introduction of a VAT typically include a simple (single or, sometimes, (Figure 4.9). dual) rate structure to facilitate administration 4.19. The implementation strategy for the reform and compliance, and a comparatively high process is critical for success. A number of mandatory registration thresholds to exempt failed VAT introductions can be explained the majority of small traders (Gnossen, 1992, by political challenges and poor planning, in International Tax Dialogue, 2005, Kloeden, particular regarding administrative capacity 2011). However, based on the extensive country constraints. Ghana, for instance, repealed a experience in adopting VAT, it has also been VAT a few months after its initial introduction noted that not all these design features are in 1995. In 1997, the VAT was successfully “attainable or even desirable in the context of a re-introduced, following more extensive particular country at a particular time” (Bird and preparation of the revenue administration Gendron, 2011). Some flexibility regarding the and business community and relying on a level of exemption thresholds, use of exemptions 43 of multiple rates may thus be required in the with the regional external tariff would lead political process to build support for a VAT. to a substantial increase in tariff revenue for Haiti (30.1-47.7 percent) while revenue from E. … And Taxes on International Trade inspection fees and excise taxes would decline as imports decline. The effect on TCA revenue Could Be Streamlined would be ambiguous due to the fact that TCA is levied on the tariff inclusive value of imports. 4.21. While Haiti is one of the most open economies, Depending on the reactiveness of imports to the trade taxes account for a substantial amount of change in tariffs, the effect could thus be slightly fiscal revenue. Trade taxes from customs duties positive or negative. The total change in revenue and inspection fees are the largest source of fiscal collected at the border is expected to increase revenue, jointly accounting for one third of total between 7.6 and 14.8 percent. However, it shows government revenue (Figure 4.1). While Haiti is that this would come at the cost of a substantial one of the most open economies to trade with increase in protection, and hence domestic prices, very low tariffs, it imposes various fees that in particular for products in the category food add up to the cost of importing goods into the and beverages (both for household consumption country. A 5 percent inspection fee is applied to and industrial inputs). This reform scenario almost all imported products increasing tariff does not include any change in the 5 percent revenues by about 75 percent in 2013. inspection fee, suggesting that while Haiti would 4.22. Ongoing discussions on tariff changes are align its tariffs with regional neighbors under unlikely to lead to significant revenue effects. the CET, actual levels of protection including the The 2013 budget was proposing a number of inspection fee would be substantially higher. tariff adjustments. The proposed tariff changes 4.24. Inspection fees could be replaced with a would have affected one fifth of total tariff lines combination of higher regular tariffs and the at 8 digit level. The revenue implications of these removal of exemptions with no loss in total tariff adjustments were simulated using the revenue. As explained in the previous paragraph, World Bank’s Tariff Reform Impact Simulation inspection fees in the magnitude applied by Haiti Tool (TRIST) (Box 4.4). The results suggest that mask the true level of protection in the country. the revenue gains from this reform would be Being applied across the board with an identical limited, equivalent to about 4 percent increase in rate for almost all products, they are unlikely tariff revenue (scenario 1 in Table 4.2 and Figure to be in line with the country’s strategic trade 4.10). Changes in the level of protection would objectives that would normally suggest placing a also be limited for most sectors, but the highest higher tariff on consumer goods and lower tariffs increase in protection by 1.3 percentage points on key inputs and capital goods. Scenarios 3 and is for capital goods, which may have detrimental 4 therefore show different illustrative examples effects on innovation and productivity growth in for replacing the inspection fees with tariffs. the longer term (Eaton and Kortum, 2001). Scenario 3 is a proportionate 75 percent increase 4.23. Alignment of Haiti’s tariffs to the current on all existing tariffs, while scenario 4 combines CARICOM Common External Tariff could a 50 percent reduction in all tariff exemptions generate substantial additional revenue, but with a proportionate 20 percent increase on all also substantially raise protection and costs existing tariffs. Both scenarios are designed to for firms and consumers. Haiti is a member of have a minimal effect on total revenue (Table CARICOM, but applies lower tariffs. As shown 4.2), but Figure 4.10 shows that their impact under scenario 2 in Table 4.2, an alignment 44 Table 4.1 : Customs Duties and Inspection Fees (2013) Import Customs duties Inspection fee value (HTG mln.) revenue hypothetical effectively hypothetical amount of share of revenue hypothetical effectively hypothetical amount of share of collected revenue applied tax weighted foregone foregone collected revenue applied tax weighted foregone foregone (HTG mln.) without rate tax rate revenue revenue (HTG mln.) without rate tax rate revenue revenue exemptions without (HTG mln.) exemptions without (HTG mln.) (HTG mln.) exmptions (HTG mln.) exmptions 11 - Food and beverages (industrial 24,529 617 674 2.5% 2.7% 57 8.4% 1.225 1.226 5.0% 5.0% 2 0.2% use) FROMENT (BLE) DUR 3,783 113 132 3.0% 3.5% 20 14.9% 189 189 5.0% 5.0% 0 0.0% GRUAUX ET SEMOULES DE 385 0 13 0.0% 3.5% 13 99.2% 19 19 5.0% 5.0% 0 0.2% FROMENT(BLE) SUCRE CHIMIQMT PUR,A L'ETAT 102 3 15 2.7% 15.0% 13 81.8% 5 5 5.0% 5.0% 0 0.0% SOLIDE YC LE SUCRE INVERTI OU (INTERVERTI) 12 - Food and Beverages 22,181 2,355 2,647 10.6% 11.9% 292 11.0% 1,102 1,109 5.0% 5.0% 7 0.7% (household consumption) MORCEAUX ET ABATS DE COQS ET 2,672 580 668 21.7% 25.0% 88 13.1% 133 134 5.0% 5.0% 1 0.4% POULES, CONGELES 45 PATE CONCENTREE DE TOMATE 64 3 13 4.5% 20.0% 10 77.3% 3 3 5.0% 5.0% 0 0.0% NON CONDITIONNEE POUR LA VENTE AU DETAIL PREPARATIONS POUR 157 25 31 16.1% 20.0% 6 19.7% 8 8 5.0% 5.0% 0 0.3% SAUCES...CONDIMENTS, ASSAISONNEMENTS. 2 - Industrial supplies not 30,426 1,489 2,110 4.9% 6.9% 621 29.4% 1,460 1,521 4.8% 5.0% 61 4.0% elsewhere specified FLACONS,BONBONNES ET 748 14 187 1.9% 25.0% 173 92.3% 37 37 5.0% 5.0% 0 0.8% ARTICLES SIMILAIRES EN MATIERES PLASTIQUES BOUTEILLES, BONBONNES, 667 27 100 4.0% 15.0% 73 73.4% 33 33 5.0% 5.0% 0 0.1% FLACONS...EN VERRE. AUTRES BOUCHONS 238 7 60 2.9% 25.0% 53 88.4% 12 12 5.0% 5.0% 0 0.2% 3 - Fuels and lubricants 35,515 879 7,121 2.5% 20.1% 6,242 87.7% 523 1,765 1.5% 5.0% 1,242 70.4% ESSENCE POUR MOTEURS 12,245 837 7,077 6.8% 57.8% 6,241 88.2% 275 612 2.2% 5.0% 337 55.0% (GAZOLINE) NAPLITE ET BENZITE. PREPARATION LUBRIFIANTE 26 0 1 1.3% 5.0% 1 74.9% 1 1 5.0% 5.0% 0 0.0% POUR MATIERES AUTRES QUE CEUX DU NO 340311. BETTER SERVICES BETTER SPENDING, Import Customs duties Inspection fee value (HTG mln.) revenue hypothetical effectively hypothetical amount of share of revenue hypothetical effectively hypothetical amount of share of collected revenue applied tax weighted foregone foregone collected revenue applied tax weighted foregone foregone (HTG mln.) without rate tax rate revenue revenue (HTG mln.) without rate tax rate revenue revenue exemptions without (HTG mln.) exemptions without (HTG mln.) (HTG mln.) exmptions (HTG mln.) exmptions HUILES MINERALES 809 40 40 5.0% 5.0% 0 0.4% 40 40 5.0% 5.0% 0 0.2% SPECIALES ET AUTRES HUILES LUBRIFIANTES COMPOSEES. 4 - Capital goods (except transport 9,976 237 308 2.4% 3.1% 71 23.1% 425 494 4.3% 5.0% 69 14.0% equipment), and parts and accessories thereof PARTIES ET ACCESSOIRES DES 115 1 6 0.7% 5.0% 5 85.8% 1 6 0.7% 5.0% 5 85.8% MACHINES DU N. 84.71 GROUPES ELECTROGENES D'UNE 104 2 5 1.9% 5.0% 3 62.8% 4 5 3.9% 5.0% 1 22.7% PUIS. >375 KVA PARTIES DES MACHINES DES 79 1 4 1.0% 5.0% 3 79.3% 1 4 1.0% 5.0% 3 79.2% NOS 85.01, 85.02 5 - Transport equipment and parts 9,393 737 831 7.8% 8.8% 94 11.4% 432 470 4.6% 5.0% 38 8.1% and accessories thereof VEH. TOUT TERRAIN , CYL. > 2500 1,114 78 111 7.0% 10.0% 33 30.0% 44 56 4.0% 5.0% 12 21.0% 46 CM3 , DIESEL , NEUFS ACCUMULATEURS AU PLOMB 203 17 41 8.6% 20.0% 23 57.0% 4 10 2.1% 5.0% 6 57.0% AUTRE QUE DU 850710 VEH.TOUT TERRAIN , CYL. >= 788 32 39 4.0% 5.0% 8 19.8% 35 39 4.4% 5.0% 4 11.2% 2200 CM3 , MOINS DE 2 TONNES , DIESEL , NEUFS 6 - Consummer goods not 12,235 1,199 1,279 9.8% 10.5% 80 6.3% 571 612 4.7% 5.0% 40 6.6% elsewhere specified IMPRIMES AUTRE QUE DU NO 103 9 21 8.3% 20.0% 12 58.5% 3 5 2.8% 5.0% 2 44.4% 491191 MEUBLES EN PLATIQUES AUTRE 61 5 12 8.6% 20.0% 7 57.1% 1 3 2.1% 5.0% 2 57.1% QUE CEUX DU NO 940370 11 OUVRAGES EN MATIERES 130 8 13 6.0% 10.0% 5 39.7% 6 6 4.8% 5.0% 0 3.1% PLASTIQUES ET MATIERES DES N.3901 A 3914 NDCA 7 - Goods not elsewhere specified 72 0 0 0.2% 0.2% 0 0.4% 0 0 0.1% 0.1% 0 0.4% Grand Total 168,922 7,513 14,970 4.4% 8.9% 7,457 49.8% 5,737 7,197 3.4% 4.3% 1,460 20.3% Source: Authors’ calculation based on raw data provided by Haiti Customs. Table 4.2 : TRIST Simulations on Four Scenarios of Tariff Reform Scenario 1 Sceénario 2 Scenario 3 Scenario 4 status quo implementing tariff reforms aligning tariffs with Remove inspection fee, Replace inspection fee, (bln. HTG) in 2014 budget law CARICOM CET (% change) increase all customs tariffs reduce all exemptions by50 (% change) by 75 % (% change) % and increase all customs tariffs by 20 % (% change) low E high E low E high E low E high E low E high E Total importations 144,3 -0.1% -0.2% -1.4% -2.6% 0.2% 0.4% 0.0% 0.1% Total recettes droits de 7,5 3.9% 3.5% 47.7% 30.1% 72.6% 70.2% 71.8% 64.3% douane 47 Total recettes frais de 5,8 -0.1% -0.3% -1.8% -3.9% -100.0% -100.0% -100.0% -100.0% vérification Total recettes droits d'accise 0,8 -0.2% -0.4% -0.3% -0.6% -0.3% -0.5% 1.0% 2.0% Total recettes TCA 10,4 0.1% -0.1% 1.3% -1.8% 4.5% 4.5% 3.0% 4.5% Recettes totales 24,5 1.2% 1.0% 14.8% 7.6% 0.7% 0.0% -0.1% -1.8% Source: Authors’ calculations, using 2013 Haiti customs data and the World Bank’s Tariff Reform Impact Simulation Tool (TRIST) with low elasticities 1.5 for substitution and 0.5 for demand, and high elasticities 5 for substitution and 1 for demand. See Box 4.3 for more information on TRIST. BETTER SERVICES BETTER SPENDING, on protection would vary significantly across consumer goods. Scenario 4 would provide for sectors. Compared to the status quo inclusive of lower tariff protection on all sectors except for inspection fees, scenario 3 would significantly fuels and lubricants, which account for the major reduce protection for industrial inputs and share of tariffs exemptions to be removed under capital goods at the expense of higher tariffs on this scenario. Figure 4.10: Applied Protection from Tari s and Inspection Fees for Four Reform Scenarios 25 20 15 10 5 0 Total Food and Food and Industrial Fuels and Capital Transport Consumer imports beverages beverages supplies lubricants goods equipment goods (industrial) (household) Tari Inspection fee from left to right bars for each sector represent: statu quo: current tari and inspection fee collection (2013 data) Scenario 1: implementing tari reforms in 2004 budget law Scenario 2: aligning tari s with CARICOM CET Scenario 3: Remove inspection fee, increase all tari s by 75% Scenario 4: Remove inspection fee, reduce all exemptions by 50% and increase all tarifs by 20% Source : Authors’ calculations, using 2013 Haiti customs data and the World Bank’s Tari Reform Impact Simulation Tool (TRIST). Box 4.4 : The World Bank’s Tariff Reform Impact Simulation Tool (TRIST) TRIST was designed with the specific task of providing policy makers with detailed insights by product and tax instrument into the short-term effects of trade reform. By its comparative static nature, TRIST allows the comparison of two states - one in which the base values of policy instruments (such as tariffs) are unchanged and another in which these base values are exogenously changed. TRIST is an Excel based tool that does not require expert knowledge in modelling or specialized software and is designed to handle very disaggregated information. An integral part of TRIST is the trade model that underlies the quantification of the effects of trade reform. It is based on five core assumptions: 48 BETTER SPENDING, BETTER SERVICES First, the model is derived from standard consumer demand theory and utilizes elasticities to determine the magnitude of the demand response to the price changes that result from a tariff reform. Second, the calculations are based on the standard Armington (1969) assumption of imperfect substitution between imports from different trading partners since consumers distinguish products by the place of production. This intuitive assumption is standard in empirical international trade work and implies that a fall in the price of imports from country A relative to country B will only lead to a partial and not complete substitution of imports from country B with imports from country A. Third, the model does not allow for direct substitution between different products. In other words, each product is modeled as a separate market and in isolation from other markets. This is perhaps the strongest assumption used in the model. However, a relaxation would not only complicate computations but would also generate a need for a range of additional ad- hoc assumptions regarding the precise design of the additional substitution effect and its parameterization. Fourth, it is assumed that all changes in tariffs are fully passed on and that the world price remains unchanged. That is to say an infinite supply elasticity is assumed of imports so that changes in demand in the importing country are small enough relative to the world so that they have no effect on the world price of the product. Fifth, the trade model in TRIST is a partial equilibrium model that treats demand for each product in isolation from the rest of the economy. Hence, it does not take into account inter- and intra-sectoral linkages or the economy wide impacts of tariff changes. But this is not the primary objective of TRIST, which is designed so as to avoid the degree of aggregation of the data that would be necessary in order to implement economy wide computable equilibrium models and to remain simple and transparent in its assumptions, with the flexibility to adjust the key parameters. A detailed discussion of the trade model in TRIST can be found in Brenton et al. (2009). As most partial and general equilibrium trade models, TRIST uses elasticities as the parameters of the model that determine how trade flows react to a given change in prices. Elasticities are notoriously difficult to estimate and so detailed and robust estimates of the required elasticities for import demand and substitutions between imports from difference trading partners are not readily available in the literature. In the absence of empirically estimated country specific elasticities the model is calibrated with different scenarios of standard elasticities for substitution between trading partners (‘exporter substitution’) and the overall effect on import demand (‘demand elasticity’). Standard elasticities, shown in TRIST result tables as “low ɛ”, are set at 1.5 for exporter substitution and 0.5 for demand substitution, whereas “high ɛ” are set at 5 for exporter substitution and 1 for demand substitution. Source: World Bank (2014) 49 50 BETTER SPENDING, BETTER SERVICES Chapter 5: Getting a greater growth dividend from public investment27 This chapter examines the public investment management system in Haiti by looking at four key steps: project identification, planning, execution, and ex-post evaluation. The regulatory framework for public investment in Haiti is complex and not consistently applied. Sectoral strategies to anchor government broader priorities into sector programs and projects are missing. Tools, staff capacity and skills to adequately evaluate and select projects are also missing, leading to the funding of projects that are neither fully assessed nor prioritized. Internal control processes and mechanisms are inadequate which, combined with an absence of project technical monitoring and financial accounting, translate into high fiduciary risks. Ex-post audit of investment projects is ineffective. Overall, the system is weak and would call for strong action if the country’s growth outlook and social indicators are to be improved. A. High Public Investment Doesn’t high in relation to Latin American countries and Low Income Countries, but has not generated a Translate into Faster Growth in commensurate acceleration of economic growth. Haiti…27 Despite the high level of capital expenditure, Haiti’s growth rate has been lagging behind, 5.1. Despite rising public investment, economic questioning the efficiency of public investment growth has not accelerated so far in Haiti. The (Figure 5.1). As discussed in Chapter 1, Haiti level of public investment expenditure in Haiti is still performs poorly for selected infrastructure indicators, such as access to electricity, roads 27 Prepared by Josue Akre, Emeline Bredy Mamadou or ports. Against this backdrop, high public Deme, Andy MacDonald, Fabienne Mroczka, and investment should have been expected to Gerard Verger. 51 Figure 5.1 : Capital Investment (a) Capital Expenditures, 2003-2013 (b) GDP Growth, 2003-2013 (Percentage of GDP) (Real GDP 2002 = 100) 180 20 160 15 140 10 120 5 100 0 80 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Haiti LAC LIC Haiti LAC LIC Sources: IMF and World Bank Sources: World Bank contribute in reducing bottlenecks to faster 5.3. The heavy reliance on donors has weakened growth. Haiti’s public investment management. As indicated in a recent IMF study, the efficiency 5.2. This paradox is, however, nothing new in Haiti. of public investments in Haiti ranks among High levels of investments have been common the lowest quartile in the sampled countries throughout the 1965-2004 period, interrupted drawn from many regions of the world (Figure only during the embargo of the first half of the 5.2) and particularly low in the early stages of 1990s, when investment decreased by around the investment cycle (Figure 5.3). Haiti’s public 4 percentage points of GDP. After this decline, investment management exhibits a number of investment climbed back to levels above 20 distinctive features and practices common to percent of GDP, attaining at some point levels countries that are aid-dependent, including- of more than 30 percent of GDP. And yet, those weak appraisal capacity and reliance on donors high levels of investment did not translate into to design good projects. These hamper the economic growth. Several reasons have been effectiveness of public investments (Box 5.1). put forward, ranging from deficiencies in the While the Government’s Strategic Development country’s national accounting system to chronic Plan (PSDH) provides broad guidance on lack of maintenance or simply the unproductive Government priorities, sectoral strategies to nature of the investment itself. Poor past donor guide the prioritization of projects are lacking. coordination and the high volatility of external This leads to a Public Investment Program (PIP) aid have also been argued to have affected the composed of projects that are neither fully impact of investments in Haiti. This chapter will assessed nor prioritized. Furthermore, there is focus on the quality of the public investment no effective ex-ante control on disbursements management system. 52 BETTER SPENDING, BETTER SERVICES Figure 5.2 : Public Management E ciency Index (0=Lowest, 4=Highest) 4 Quartile 1 Quartile 2 Quartile 3 Quartile 4 3 2 1 0 Belize Congo Solomon Islands West Bank/ Gaza Yemen Lao Sao Tome & Princ. Gambia Burundi Togo Senegal Gabon Chad Sierra Leone Sudan Haiti Swaziland Trinidad & Tob. Guinea Nigeria Barbados Djibouti Tanzania Kyrgy Rep. Égypt Uganda Indonesia Kenya Azerbaijan Benin Cambodia Pakistan Mozambique Albania Montenegro Ethiopia Mauritania Jamaica Mongolia Kosovo El Salvador Namibia Malawi Philippines Ghana Cote d'Ivoire Zambia Turkey Lesotho Macedonia Ukraine Madagascar Serbia Bangladesh Belarus Burkina Faso Afghanistan Mali Jordan Rwanda Moldova Botswana Kazakhstan Armenia Bolivia Peru Thailand Tunisia Colombia Brazil South Africa Appraisal Selection Managing Evaluation Source : IMF Figure 5.3 : Public Investment Management Index – PIM (0=Lowest, 4=Highest) 1.3 Evaluation 1.7 Managing 1.2 Selection 0.0 Appraisal 0.0 0.5 1.0 1.5 2.0 2.5 Haiti Low Income Countries Latina America and Caribbean Source : IMF based on projects’ physical progress against of scarce public resources. Finally, even though plans. More importantly, domestically the existing legal framework is acceptable funded capital expenditures are not properly for the management of public investments, accounted for, tracked and reported, creating an its requirements are rarely respected, with environment conducive to a lack of transparency numerous processes and procedures that, when and accountability, as well as to mismanagement not redundant, are excessively elaborate. 53 Box 5.1 : Common Features of PIM in Donor-Dependent Countries PIM systems in donor-dependent settings tend to exhibit the following distinctive features: Investment guidance, project development and preliminary screening: government strategy documents, such as a PRSP, tend to be directed towards the donors, rather than covering both external and domestic investment in an integrated and coherent manner. They are at a level of generality that limits the extent to which they can provide a basis for preliminary screening of projects, and are often not supported by effective sector strategies. Formal project appraisal: there is a reliance on donors to conduct appraisal, with a serious lack of appraisal capacity within government; and a lack of guidance on defining the project preparation process and on how to appraise domestically-financed projects and PPPs. Donor capacity-building on appraisal tends to be agency-specific, with little or no domestic training capacity. Independent review of appraisal: reflecting reliance on donors, there is a lack of capacity for independent review, either of donor projects or domestically-financed projects. Project selection and budgeting: the budget is divided into a recurrent and a development budget, with weak integration between the two and substantial off-budget aid. The use of PIPs remains quite common, but these can be poorly connected to fiscal policy and the budget. In practice, a PIP tends to be more a coordination tool than a tool to manage the project portfolio strategically or to help enforce review of individual project proposals before they can be considered for budget funding. Agreement by a donor to finance a project is tantamount to the project being included in the budget – subject to basic screening for consistency with a PRSP (which is not difficult given their generality) and any required counterpart financing being affordable. Project implementation: unpredictability of donor funding (especially budget support) interrupts project implementation due to lack of alternative financing. Weak project management capacity induces donors to set up multiple Project Implementation Units (PIUs) within implementing agencies that initially help to speed implementation and compliance with fiduciary standards, but which cut across and impact negatively on in-line capacities and accounting and reporting systems. Procurement is undertaken by PIUs or donors to donor standards rather than national procurement standards. Project adjustment: reliance on donors to trigger review of any projects that are off-track. No similar mechanism for domestically-financed projects. Facility operation: formal hand-over procedures on completion of donor projects, but inadequate asset registration systems; and inadequate funding for operations and maintenance, in part due to weak integration of recurrent costs of donor projects in fiscal policy and budgets. Basic completion review and evaluation: reliance on donors to undertake reviews and evaluations of their projects. Otherwise, little or no systematic basic post-project review, let alone evaluation, and little systematic use made of findings from donor evaluations to improve future project design and implementation. Source: Rajaram et al, (2010) 54 BETTER SPENDING, BETTER SERVICES Table 5.1 : Set of Projects Reviewed by Ministerial Affiliations Number of projects Ministries selected 1 Plan 4 2 Health 3 3 Agriculture 2 4 Finance 2 5 Infrastructure 2 6 Education 2 7 Justice 1 8 Interior 1 9 Culture and Communication 1 10 Environment 1 11 Sport 1 Total 20 Figure 5.4 : Four Stages of the Project Life Cycle Identi cation Planning Execution Evaluation Consistent with Ongoing projects Monitoring by MPCE (DPES, PSDH Operational Plan executing agency DIP, DSE) Consistent with Financial Plan External Sector Ministry Sectorial Strategy FIOP monitoring (MPCE, MEF Included in MEF, Sector CSCCA Projects in ministries, Triennial preparation Donnors Investment Plan donors) Feasibility Study Quartely Reports Project File (sector ministry submits Socio- economic Annual balance to its UEP and study sheet of physical MPCE for analysis) Project progress and Document (DDP) nancial FIOP statement from Inclusion in PIP beginning of with above project Documentation Source: Haitian authorities 55 Box 5.2 : Haiti Legislative Framework Haiti, Constitution, 1987. Finance Law Decree, February 16, 2005 on preparation and execution of finance laws. July 2001 Regulation setting the State’s expenses budgetary nomenclature. November 2001 Regulation setting the State’s resources budgetary nomenclature. Arrêté, February 16, 2005 setting general regulation on public accounting. Regulation, October 2006 setting the State’s chart of accounts. Law N° CI. 06 2009 009 establishing general rules related to procurement and concession agreements for work and public goods. Decree, October 4, 1984 setting up the Public Investment Fund within the Ministry of Planning. Arrêté, September 17, 1985 defining application conditions of the Public Investment Fund. Arrêté September 5, 2009 establishing public procurement thresholds and the National Procurement Commission’s intervention thresholds. Arrêté October 26, 2009 determining application conditions of general rules on public procurement and concession agreements work of public service. Arrêté October 26, 2009 establishing procedures for the organization and functioning of the CNMP, the National Procurement Commission. Arrêté December 19, 2005 establishing special rules for the statute of Inspectors of IGF, Treasury Public Accountants and Budget Financial Controllers. Decree, March 17, 2006 creating the Ministry of Economy and Finance decentralized technical service called the General Inspectorate of Finance. Decree May 17, 2005 revising the Civil Service Statute. Decree November 23, 2005 establishing the organization and functioning of the SAI. Source: http://www.sdn.mefhaiti.gouv.ht/sdn_lois.php#A2005 5.4. The chapter structure follows the project cycle B. … Because of an Overly Elaborate from planning to completion (Figure 9.4). and Rarely Respected Regulatory The analysis draws among other sources on (i) interviews with key PIM stakeholders, and (ii) Framework a representative sample of 20 recent investment 5.5. Haiti’s public financial management system has projects (Table 5.1 and Figure 5.4).28 The sample not significantly evolved since 2008. Reports focused on the eleven ministries with the largest from various sources (World Bank 2008,29 investment budgets (76 percent of the total PEFA 2011,30 IMF 2013,31 and the Directorate of number of projects in the 2012-13 investment Monitoring and Evaluation of Haiti’s Ministry budget, included projects from all four sources of investment funding - Treasury, Petrocaribe, bilateral and shared donor financing, and 29 Haiti Public Expenditure Management and Financial represented all Government sectors. Accountability, World Bank 2008. 30 PEFA, PEFA Evaluation Haïti, January 2012. 31 IMF Country Report No. 13/91, Haïti. Selected Issues, 28 Total population of approximately 700-800 projects. March 2013. 56 BETTER SPENDING, BETTER SERVICES of Planning32) describe the same shortcomings 5.7. As a result, the formal requirements are rarely across the spectrum of Public Financial respected. Failure to adhere to procedures Management (PFM) components, with few and controls, as designed in the regulatory marginal improvements. Box 5.2 outlines the framework, is not sanctioned. The requirements PFM legislative framework. are hence largely nominal in practice, not consistently applied or enforced. Our sample 5.6. Even though the existing legal framework indicates the lack of compliance with the is acceptable for the management of existing regulatory framework throughout the public investments, its requirements are project lifecycle. For example, the review showed overly elaborate for a capacity-challenged that first level screening seldom takes place; environment. The Haitian legal framework thus calling into question whether projects includes sound laws, decrees, arrêtés, and meet minimum criteria of consistency with the regulations covering all aspects of preparation, strategic goal of government or meet budget presentation, and control of budget execution classification tests for inclusion as a project related to financial transactions. The public rather than as a recurrent spending item. This investment management system more is a critical step in a well-functioning public particularly is governed by a Decree and an investment management system to ensure that Arrêté.33 The Decree of October, 4 1984 created limited resources are not wasted and contribute the Public Investment Fund within Ministry of to the Government’s policy agenda. Similarly, Planning and External Cooperation (MPCE) from our sample of 20 projects, only 6 (30 and the Arrêté of September 17, 1985 set out its percent) of these had all of the required project execution. Both documents include information documentation to be included in the PIP (e.g. guiding the operation of the PIM system in Project Document, supporting studies, financial Haiti. In addition, MPCE has recently completed plans, FIOP). A quarter of the projects did not a procedures manual governing the preparation even have FIOPs (Identification and Project and execution of investment projects in Haiti.34 Operation File), although the existing regulatory Regulation establishes the requirements for the framework requires projects to prepare one to identification, preparation and execution of receive disbursements of the allocated credits. investment projects. This framework requires a great number of processes and procedures that, when not redundant, are overly elaborate and C. … Little Strategic Guidance to prescriptive. Select Projects 5.8. While the Strategic Development Plan provides broad guidance on Government priorities, 32 Diagnostic du système de suivi et d’évaluation programme et projets d’investissement public, janvier, many sectors and ministries remain without MPCE/DES, 2014. a sector strategy that translates this document 33 Decree of October 4, 1984 creating the Public into sector priorities. Broad strategic guidance Investment Fund within the Ministry of Planning; and for public investment is an important starting Arrêté of September 17, 1985, stating the execution of point to anchor decisions and to guide sector- the October 4, 1984 decree. level decision-makers. Strategic guidance 34 Manuel des Procédures sur la Gestion de l’Investissement Public Tome I: Élaboration du Programme ensures that investments are chosen on the basis d’Investissement Public, Tome II: Exécution, Suivi et of development policy priorities. As mentioned Évaluation du Programme d’Investissement Public, in Chapter 3, the Government of Haiti completed 2014 57 in May 2012 a Strategic Development Plan financial and social assessments and prepare (PSDH), detailing the country’s vision to become Project Documents. an emerging economy by 2030. This vision is, • A lack of clear guidance on the criteria to be however, too broad to serve as a basis for project employed as to when, and how pre-feasibility, selection. While some sectors have strategies, feasibility and socio-economic studies should there is again not enough information in terms be undertaken. of timing and selection criteria that could assist ministries decide which project to include in the 5.11. As a result, the selected projects are often PIP. As a result, links between the Government’s not fully mature, leading to delays in their investment projects, its national policies and implementation. The lack of rigor in project its underlying strategies are tenuous. From analysis, justification and selection leads to our sample of 20 projects, only a quarter were immature projects being included in the budget supported by a sector strategic plan. which may fail to be successfully implemented or even begun. Consequently, the PIP does not 5.9. Against this backdrop, ministries face represent a strong government commitment difficulties in setting priorities. The regulatory to perform. This contributes to delays in the framework in Haiti requires ministries to execution of projects. prepare a list of potential projects that are then to be subjected to feasibility studies on the basis 5.12. Alterations to projects once approved are of key operational, financial, socio-economic frequent. This top-down approach can preclude and environmental factors to avoid selecting ministries from executing activities they consider wasteful or unfeasible projects (Figure 5.5) priorities through the normal PIP channel and provides a description of the required processes encourage them to find alternative resources to for registering a project in the PIP). In reality, execute these activities, or (with more adverse however, the process is seldom followed. Since consequences) to change the objectives of many ministries lack a formal strategy document projects after they have been approved in PIP. that establishes policy priorities and associated Moreover, the MPCE can and does arbitrarily projects, the MPCE selects frequently the replace PIP projects, without transparency or projects that will appear in the PIP, sometimes stated justification, thus reinforcing the top- on the basis of political considerations. In the down character of actual prioritization. In the sample of 20 projects, only 7 had been prioritized sample of 20 projects, 4 were not maintained in for PIP inclusion by the responsible ministry. their original, approved PIP form. 5.10. Consequently, projects included in the PIP are 5.13. The PIP includes recurrent or operational often only at the concept stage, and have not expenses. Recurrent costs are often not been fully developed. Basic documents such as considered in the operating budget when the Project Document, feasibility studies, and establishing the investment project. The lack of economic, financial and social assessments are a multi-year expenditure plan doesn’t provide often not prepared and submitted to the MPCE. incentives for increasing operating costs as Nonetheless the project is included in the PIP (to a result of completion of investment project. ensure its inclusion in the budgetary document) As a result, the PIP includes non-investment because of: spending and over-estimates the amount of public investment taking place in the country, • A lack of capacity and resources to conduct distorting measurement of its impact on and analyze feasibility studies, and economic, economic growth following implementation. 58 BETTER SPENDING, BETTER SERVICES Figure 5.5 : Required Process for Adding New Projects to the PIP National PSDH priorities MPCE Sector Sector to validate strategies priorities allegnment with PSDH The Minister must validate the Projects Ranking allignment with the sector strategy Projects Update with project Projects Projects near execution information approuved underway completion Priorisation by ministries Ongoing, approved projects that are in early stage, with strong allignment with the sector strategy Three years Medium-Term (MPCE to validate) Investment Plan (PTI) Forecasts FIOP Contraints Public Investment Budgetary Plan (PIP) envelopes Source: Haitian authorities 59 In our sample of 20 projects, almost a third performed by the MPCE without reference to was operating budget projects. These were the sponsoring ministry. generally related to the provision of subsidies to 5.16. Despite the Triennial Investment Plan, true commodities, individuals, and industries. multi-year planning is still lacking. While a 5.14. The PIP, and by default the National Budget, three-year investment plan has been prepared often does not include all donor-funded for 2014-16 reflecting the first three years of projects. This results in an undervalued PIP and the PSDH, this tool is not used as a multi-year the potential for a lack of convergence between planning tool. It is not anchored in available the objectives of the externally-financed projects resources, and the procedures to make changes and the objectives of the government and to the PTI and its monitoring are still unclear. sectors. This is counter to the spirit of the Paris In addition, the three-year plan is not sufficiently Declaration, specifically on the use of in-country grounded on ongoing programs and projects, systems wherever feasible. This also means that disconnecting the PTI and the PIP. Against this the related project costs are not included in the backdrop, some projects that have already been PIP, are not reported in the Haitian national selected, or even under implementation, may budget, are not fully monitored as external be abandoned or deferred. The review found, debt and are not controlled by government for instance, some examples of abandoned or financial systems. Finally, the impact on the deferred projects in the health sector. operating budget of completed projects funded 5.17. Strategic coordination among ministries for by externally-funded projects is often neglected. multi-sectoral projects is limited. The MPCE Our discussions revealed, for example, that tends to treat such projects as a number of the operating costs of an externally-financed independent projects, with little, if any, ongoing hospital had not been taken into consideration, project coordination. For example, delays in impacting its entry in operation once completed the construction of basic infrastructures, such (Box 5.3). as roads, has been experienced during the construction of a new health center, jeopardizing D. … A Budgeting Process That is Not the effective use of the hospital once built. Followed This contributes to higher risks for project completion and a potential adverse impact on 5.15. While there is a well-defined annual budget total project outlays. Figure 5.6 illustrates the cycle for investments, it is seldom followed. The budget formulation process for the preparation process for the preparation and the execution of of the investment budget. the operating budget and the investment budget 5.18. Furthermore, project classification used in are established by decree, according to a set the budget and in the PIP is unreliable. The schedule. These two budgets are prepared in same PIP reference can be used for more than parallel through separate processes managed by one project. It is not uncommon for some the MEF and the MPCE, respectively. However, projects to be assigned different codes from in reality the set schedule is seldom followed one year to another, or for identical codes to be and important delays have been observed. In assigned to different programs or projects. This addition, the current budget cycle does not means that the reporting of the budget vs. actual permit appropriate ministry participation, expenditures cannot be readily determined; such particularly when it comes to project selection. reporting is the foundation of accountability for At times, selection (deletions, additions) is 60 BETTER SPENDING, BETTER SERVICES Box 5.3 : The Share of Recurrent Expenditures in Investment Projects Investment programs and project are not reported by economic classification, making it difficult to identify the recurrent share of the investment budget. However, some investment programs are included in SYSDEP with their respective economic classification (in 2012-13 approximately 50 percent of PIP in monetary terms was included in SYSDEP). The Figure below depicts the four top spending categories for investment programs recorded in SYSDEP. It is evident from the Figure that wages and salaries account for the largest share of spending across the observed period. Top Spending Categories for Investment Programs in SYSDEP 2009-13 (percent, budgeted) 80 60 40 20 0 2009 2010 2011 2012 2013 Wages and salaries Purchase of goods Services and other fees Immobilization Source : Haitian authorities Moreover, looking at the nature of investment projects in the PIP, certain types of “recurrent” projects can be identified, such as projects with a lifecycle that extends beyond 4 years or projects with an “institutional” or “appui” label. The Table below shows the amounts budgeted for “recurrent” projects, compared to the total amount of treasury funded projects in that same year. Budgeted Amounts for “Recurrent” Projects Selected Budgeted Treasury funded % projects (prevision) projects 2007 879,019,890 6,205,872,899 14.2 2008 855,445,146 4,000,000,000 21.4 2009 728,362,595 4,360,191,490 16.7 2010 2,603,714,685 11,300,000,000 23 2011 2,432,000,000 12,501,500,000 19.5 2012 2,636,072,697 15,397,886,784 17.1 2013 3,000,148,456 18,011,105,415 16.7 Source: World Bank Estimates 61 Figure 5.6 : Illustrates the Budget Formulation Process for the Preparation of the Investment Budget. Source: Haitian authorities 62 BETTER SPENDING, BETTER SERVICES all contracts are still selected through single Figure 5.7 : Contracting Methods, 2012-13 sourcing (Figure 5.7). Furthermore, while the (Number of Contracts Reviewed by CNMP) percentage of competitive contracting has increased, the quality of competitive bidding is 80 still low. This limited competition can represent 70 additional costs for the budget to carry out 60 public investments. 50 40 5.21. Procurement planning is insufficient and there 30 is limited compliance with legal provisions. 20 A review of our sample indicates that only five 10 percent of the projects reviewed were included 0 2012 2013 in a procurement plan. More generally, the legal and regulatory framework mandates the Direct contracting single source selection Limited competition ministerial procurement units to publish and Open competition transmit their procurement plans for acquisitions above the threshold for competitive tenders to Source : Haitian authorities the National Procurement Commission.35 At the same time, Decree October 26, 2009, Article 220 requires departments and agencies to submit the the use of public resources, including investment quarterly reports on all procurements, regardless resources. of the threshold. However, for 2012-13, only three ministries and five public institutions had E. … Weak Procurement… submitted their procurement plans and none their quarterly reports. 5.19. Ministries lack sufficient capacity to perform 5.22. The validation of transactions at each stage of their procurement responsibilities. The the process introduces further delays. Figure study noted a particular lack of resources and 5.9 shows the various steps needed to validate, expertise of staff involved in the preparation of control, and execute procurement transactions tender documents, launching public tenders, as prescribed by law.36 In total, it should take no and their evaluation – all of which have seriously more than about four months to process a request slowed the procurement process. This is one for goods and services. However, the actual of the factors that have fostered the creation of processing time regularly exceeds the legally parallel structures outside of the official public required interval. The discussions revealed that procurement process. Except in the ministry at the earliest disbursements for new projects of agriculture, where a procurement unit is occur at the beginning of the second half of the operational, ministries do not currently have fiscal year. Indeed, the procurement process is procurement units whose main function would slowed down at each step by an administrative be to facilitate and ensure more transparency in procedure that follows every hierarchical level in the procurement process. 5.20. A large share of contracts are awarded using 35 Article 66 of the Decree of 26 October 2009 and single source selection, resulting in limited Article 5 para . 4 of the Act competition. Even though there has been 36 Loi fixant les règles générales relatives aux marchés publics et aux conventions de concession d’ouvrage de some progress over the last year, about half of service public, CL 06 2009-09. 63 Table 5.2 : Procurement Thresholds for State institutions (in millions of Haitian Gourdes)41 Procurement Process Works Goods and service Professional services Memorandum Invoice .. Amount<8 .. Direct procurement 8 ≤ Amount < 40 8 ≤ Amount < 25 8 ≤ Amount < 20 Public tender Amount ≥ 40 Amount ≥ 25 Amount ≥ 20 the organization, where every issue is submitted 5.24. Public procurement thresholds are too high to the top for resolution before the authorization for the levels of budget expenditure. Under the flows back down to its point of origin. current regime, a large proportion of purchases of goods and services are excluded from 5.23. Stakeholders have incompatible competitive tender by the National Procurement responsibilities in the public procurement Commission or under the provisions of the Law system. The study identified a number of on Public Procurement.40 Procurements under instances where procurement stakeholders had HTG 8 million (approximately USD180,000) multiple and conflicting roles. Examples include: are not covered by the law; such procurement • The National Procurement Commission is can be implemented with a memorandum both a regulatory body that carries out ex- invoice by the ministry. Between 8 million HTG ante validation controls over procurements, and the three limits of 20, 25, and 40 million as well as ex-post financial investigations and HTG for procurements of works, goods and independent audits, while at the same time services, and professional services respectively, advising in case of disputes.37 procurements are regulated by the law on Public • The SAI, whose main role is to perform an ex- Procurement. Above these three respective post audit of the government’s accounts and thresholds, procurements are regulated by law to verify stated assertions must also perform and controlled by the National Procurement the ex-ante function that “give(s) its reasoned Commission. Table 5.2 presents the details. 41 opinion on all draft contracts opinions, 5.25. There are cases of misuse of the emergency agreements and conventions to financial, provisions for direct contracting. The Law commercial or industrial which the State is (Article 7) and Decree relating to a declared party”.38 State of Emergency allows the government to • At the ministerial level, the Ministerial suspend all ex-ante controls on the release of Procurement Committees prepare Biding funds, accounting for expenses to undertake Documents and its members participate in projects, budget allocations, procurement of the Committee of bid opening and evaluation contracts and the SAI ex-ante visa requirements of tenders.39 for procurements (Figure 5.8), which can lead to some undesirable results (increase use of direct contracting, absence of document review 37 Loi fixant les règles générales relatives aux marchés publics et aux conventions de concession d’ouvrage de from the CNMP, and observed deficiencies service public (CL Act,) Article 10, Juin 2009. 38 Decree, Establishing the organization and functioning of the Superior Court of Auditors and Administrative Disputes, Article 5, le 23 November 2005 40 Ibid. 39 Arrêté, Révisant les seuils de passation de marchés 41 Arrêté, Réglementation des Marchés Publics, # 12, le publics #.117, le 4 décembre 2006 14 Février, 2005. 64 BETTER SPENDING, BETTER SERVICES Box 5.4 : INTOSAI Conditions Supporting Corruption • Excessive concentration or centralization of authority • Complex financial administration with multiple layers of bureaucracy • Shared or non-existent accountability within the processes • Weaknesses in organizational and administrative systems • Inadequate internal controls and internal audit mechanisms • Inadequate external audit mechanisms • A lack of transparency in PFM and inadequate reporting • Inadequate salaries for civil servants Source: INTOSAI in contracts).42 Our discussions revealed overall view of the use of state resources and the that, although this procedures ought to be government’s public policy objectives. exceptional, there is evidence of overuse with a 5.27. Executing entities often operate separately number of cases using emergency procedures in from ministries. Executing entities have been put areas where no emergency had occurred. in place at the behest of donors to compensate for the inefficiency of internal government services. F. … Fragmented Project Execution Although these entities have more flexibility and Administration… in project planning and implementation, and potentially can deliver implementation results 5.26. Although projects financed from domestic more swiftly, they entail several potential resources are generally included in the PIP, downsides. First, there is potentially weaker they can follow multiple, different processes linkage with the government’s overall investment and procedures (Figure 5.9). Projects financed strategy when there are significant investments from domestic resources include those financed being managed outside the government system. by the Treasury, the National Education Fund, Second, because these donor-supported entities the Roads Maintenance Fund, the Petrocaribe typically provide higher levels of remuneration, Fund, and the Debt Cancellation Fund. While they often skim off the most skilled ministerial all these funds manage the budget’s own staff, draining these scarce resources from resources, each of them has a different budget departments. In addition, public expenditure execution process. This affects how projects are information within government is less accessible selected, implemented, managed and funded and comprehensive, understating the levels and (disbursements and accounting). These different types of capital investments being undertaken processes and procedures hamper both a clear in the country. But the main drawback of these separate executing agencies is that they reduce incentives for the government and the donors to 42 Loi sur l’état d’urgence of 15 April 2010 and related Decree of September 9, 2008. 65 Figure 5.8 : Complexity in Procurement Processes 66 Source: Haitian authorities Figure 5.9 : PIP Project Execution Processes 67 Source: Haitian authorities BETTER SERVICES BETTER SPENDING, carry out the institutional reforms necessary to issuance of a visa, there is no verification make public services more efficient. required or conducted at many of these stages. 5.28. The disbursement process for the payment • In some cases the disbursement process of the quarterly instalments is extremely is not followed. In 2012, 20 percent of the cumbersome and highly centralized (Annex 1 disbursements were done directly from and Annex 2). All payment requests, regardless the MEF without following the traditional of their importance, are uniformly processed procedures requiring the review of the request as ordinary Ministry mail. For example, by the MPCE. before reaching the functional step required, 5.30. Procedures at the line ministry are equally all requests are submitted through the mail cumbersome. At a minimum, the origination receiving services, the Assistant Director, the of the request, its commitment, payment and Director General (sometimes several times), liquidation of all investment project expenses and the Minister before the requested action take place at the level of the line ministry. can be sanctioned. At various stages, the request Cumbersome procedures hamper these is accompanied by letters of transmittal, as operations. Within ministries, a request for applicable, signed by the Directors General or funds issued by the technical services may Ministers. INTOSAI has identified “complex require as many as twelve steps before the check financial administration with multiple layers of is delivered to the beneficiary. Again, multiple bureaucracy” as one of the process characteristics steps do not equate with improved internal that pose high fiduciary risks to the operations of financial controls. any government (Box 5.4).43 5.31. Sector ministries do not have an effective 5.29. Despite or perhaps because of its complexity, tool for monitoring their investment budget this process lacks effective internal controls availability or type of expenditures. The only and in some cases is not followed. For example, real indicator of budgetary availability is the the team noted: balance in the current account opened for each • A lack of periodic (for example, quarterly) project.44 Absent an effective tool to monitor field visits to verify physical and financial cash balances of individual projects, ministries progress. will have to make allocations among the different projects by using a separate set of books that are • Compliance checks of expenditure are made outside of the formal financial system. on the basis of photocopied documents, rather than originals, eliminating the effectiveness of 5.32. Payments are made on the basis of pro forma the controls. invoices and therefore before the liquidation phase. Supporting documentation to justify • The mere fact that a request is routed through, receipt of the goods or service is not attached and a visa is given by, a director, the deputy with the final bill, even retrospectively, to the director, the head of department, and the request justifying the payment. It is only since project manager for processing does not the creation of the Public Accountants that the strengthen the quality of controls. This is service rendered will be justified on an ex-ante because, contrary to the prerequisite for the basis. 44 The SYSDEP system is intended for use for 43 INTOSAI, Assessment of Conditions Encouraging management of the operating expenditures in the Corruption, Vienna workshop, April 2003. operating budget. 68 BETTER SPENDING, BETTER SERVICES 5.33. The establishment of public accountants budget execution. Gathering the necessary is expected to improve the monitoring and documents to ensure proper use of public reporting of accounts. Since the Decree of funds is part of the accountant’s obligations. February 16, 2005 establishing the public In this context, public accountants that work accountant function, the MEF has been putting at the payment stage require two proofs: the in place a network of public accountants, in many supplier’s final invoice and proof of delivery cases in line ministries. The Decree reaffirms bearing the stamp of the recipient of goods the principle of segregation of duties and good and services. governance by stipulating the role of the public 5.35. The creation of financial comptrollers is also accountant and the “ordonnateur”. With the very promising. In addition to the network of implementation of the Single Treasury Account, public accountants, there is also a network of MEF is now in the process of centralizing financial comptrollers reporting to the MEF’s this network of public accountants. Their Budget Directorate whose responsibility includes responsibilities for capital expenditure are to the control and verification of budget availability. control, support, pay and account for investment The 30 financial comptrollers have, however, not expenditures and maintain financial records. yet been allocated to all line ministries. In their 5.34. The establishment of Public Accountants Units absence, line ministry staff must obtain their constitutes a significant advance. The Public payments visa from the MEF, lengthening the Accountants Units, if properly implemented, payment cycle. But the main drawback is the should help in addressing major ex-ante control separation of financial control from the ministry. issues through: Effective control requires an intimate knowledge of the underlying business documents that are • Recognition of more disaggregated verified in the process. investments: the public accountant could disaggregate actual expenditure charged to subaccount 3900 - Expenditure of projects G. … Limited Physical and Financial by project. In order not to overload the Monitoring… accounting function, these more detailed entries should be strictly limited to what is 5.36. Accounting for investment projects is limited. essential (i.e., by project and by expenditure An entry by Treasury shows only the transfer of type such as works, goods, services). This funds to the ministry’s single investment account temporary practice should cease as soon as (starting in 2014) or the project’s account at sector ministries have the capacity to keep the Central Bank. Ministries do not have an their own detailed records of project funds. accounting system for monitoring expenditures by nature; and even when they use the existing • Elimination of prepayment of services, SYSCOMPTE system (designed for another with payment only when services have been purpose), the corresponding entries are not actually rendered or when goods have been included in the overall General Ledger of the delivered. This is the worldwide practice in State. This practice greatly limits control over the public and private sectors. There is sound public investment expenditure and poses a high justification for this change, as it significantly fiduciary risk. Ministries’ opening balances are improves financial spending controls. Under high while reported expenditures in any given this change, which the Government is now in year are low (Figure 5.10). Cash balances at the process of testing, the public accountant the BRH are not increasing, however. The low is responsible for ensuring compliance with 69 • Fairness, which implies completeness, Figure 5.10: Treasury funded Public consistency and accuracy of the financial Investment 2012-13, (In Billions of Haitian Gourdes) information provided. 16 15.4 5.38. There is no integrated computerized management system for public finances or 12 independent accounting system for public investments. There are more than 25 free- 8 standing applications that respond to specific 6.5 needs, but none of them caters to the project 4.3 4 budgetary, financial and accounting management 2.6 needs. In addition, where there is a need for 0.3 0.3 additional functionality not provided by these 0 2012 2013 stand-alone systems, requirements are met by the use of Excel spread sheets. This is not simply a Opening Balance Expenditures “process” shortcoming: the lack of an integrated Available balances in project accounts (BRH) information system has significant implications Source : Haitian authorities for the oversight of public resources, because it effectively undermines any audit trail and there is no security control over changes made in Excel (Table 5.3). levels of available cash in the project accounts at the BRH suggest that actual (but not reported) 5.39. Treasury cash management does not support spending levels are much higher. a smooth, uninterrupted implementation cycle for capital projects. There is no process 5.37. This practice does not allow for meaningful in place that allows Treasury and the ministries control and reporting on PIP activities. to project their financing needs, and thus, This practice falls short of generally accepted prioritize their transfer requests to allow for a accounting principles and, perhaps more smooth project implementation. As a result, importantly does not allow for meaningful there are cash shortfalls that result in rationing, control and reporting on PIP activities, as there reducing the efficient implementation of all cannot be any preparation of consolidated capital projects the financing of which is delayed. financial reporting for investment projects. A In 2011-13, national treasury funded projects number of basic principles of public accounting received approximately 60 to 70 percent of are not met: their appropriated credits. In addition, about 80 • Unity, which requires visibility of the budget percent of disbursements for national treasury and effective financial control. funded projects were received during the last 4 months of the fiscal year in 2012, resulting in • Completeness, which requires the recognition projects not having sufficient time to execute of all charges. their activities as planned (Figure 5.11). A • Comprehensiveness, i.e. the provision of similar trend can be observed in 2013, with a sufficient information to enable effective spike in February due to post Hurricane Sandy control over budget execution. emergency activities, which led to a national budget increase. 70 BETTER SPENDING, BETTER SERVICES Table 5.3 : Summary of Applications Relating to the Expenditures and Investment Chain Name Description Principal limitation Degree of Integration ELABUD Budget development Intended only for the preparation of Semi-automatic linkage to SYSDEP; the operating budget requires the intervention of supplier SYSDEP Expenditure management, Financial Only freezes budget allocation after Registers transaction in the General controls of requisitions the visa of the expense by Public Ledger (GL) (Automatic posting to Accountant GL) Only supports expenditure directly paid by The treasury SYSCOMPTE Treasury and current accounts Excludes information on budget No interface with General Ledger management system availability Development application; not yet stabilizad SYSPIP Financial monitoring of public Application appears to have been n.a. investments abandoned GL/TBMS Geenral Ledger / Module of None No interface with SYSCOMPTE and TBMS (Integrates financial and CHEK administrative management System) CHEK Check printing and control None No GL interface SYSGEP Monitoring System for Public Under development Not interfaced with ELABUD or GL. Investment and for Projects financed Application appears to have been by external aid abandoned MGAE External aid management program At present, covers only donors No interface with the expenditure for donors and NGOs system 5.40. Monitoring of projects is sparse. In principle, Figure 5.11: Cash Flows National Treasury Funded monitoring of projects appears to be in place. Projects 2012-13 (In Percent of Credits) Quarterly project monitoring reports, as well as annual reports (bilan), are required to 70 be submitted to the MPCE by line Ministry 60 Planning Units (UEPs) and the MPCE should 50 review this information. However, these 40 reports are seldom produced and presented in 30 a timely manner. For 2012, approximately 37 20 percent of annual reports were submitted to the 10 MPCE (for 247 projects financed by treasury, 0 only 92 annual reports were received). During October November December January February March April May June August September July that period, certain ministries such as MPCE, MCFDF, MHAVE and MAE did not submit the required annual reports, which did not prevent 2012 2013 them from getting disbursements). The fiscal Source : Haitian authorities year 2013 did not fare better, with only 52 annual reports received by the MPCE for 252 projects financed by treasury (approximately 20 percent). 71 Even when quarterly or annual monitoring fixed assets that are part of the assets of the state.45 reports are submitted, the quality is questionable At present, however, its 20 auditors are limited with often no information available on verifiable as to what they can be reasonably expected to implementation indicators. For 2012, a total of achieve. They do not prepare a risk-based audit 542 activities were programmed for the 92 annual plan, and to date have not conducted any capital reports received. Information, however, was only project-related audits, despite the significant available for 170 activities (or 31 percent of total amounts mobilized in some of the investments. activities) and no information was available for Their focus has been on system-based audits, 301 activities (or 56 percent). rather than transaction-based compliance audits that are required for the public investment 5.41. Both at the central (MPCE) and the line budget and system. ministries levels, physical monitoring of projects is insufficient. From the discussions, 5.43. The development of ministry-based internal it appeared that the MPCE had not performed audit units is unlikely in the near future. any physical supervision on-site visit in the last There is a long-term vision of achieving this five years. Because of lack of budget funds to goal, but it remains to be proposed, approved, perform the work, their technical teams rarely and implemented. In any case, it represents go into the field to inspect projects, trusting a decade-long development period for full instead the progress reports of the departmental implementation; as such, it does not represent directorates. However, these reports are not a viable medium-term option. Meanwhile, IGF treated as a priority and their production is too should be used as a technical and institutional irregular for use as a basis for monitoring. In resource for line ministries and agencies’ internal fact, some institutions have even been unable audit needs and requirements. to provide any follow-up report of projects in 5.44. The external audit function in Haiti is the PIP. Against this backdrop, the Government performed by the Supreme Audit Institution cannot obtain progress reports, the exact (SAI), the Superior Court of Auditors and situation of a project, or the challenges it faces Administrative Disputes. The organic law states in order to take timely corrective actions or that: “the CSCCA is an independent institution simply report the status. Under these conditions, whose mission is to judge the actions of the it is also very difficult to assess the quality of the Public Administration, Accounts Officers and reports produced by contractors. Public Accountants and assist the Parliament and the Executive in controlling the execution of H. … And a Lack of Ex-post the laws and regulations related to Budget and Evaluation (Project Auditing and Public Accounting.46 Evaluation) 5.45. The SAI so far plays a limited role in investment budget execution. As part of the 5.42. Created in 2008, IGF is still a nascent internal development of its opinion on the government’s audit entity. IGF has the responsibility, among financial reports, it examines the public other things, to periodically audit: (i) the system for public investments, in particular the management of investment funds, programs and 45 Décret: Inspection Générale des Finances (IGF), No 47 of 25 May 2006, Article 5. investment projects, independent of the sources 46 Décret: Etablissant l’organisation et le fonctionnement of financing; and (ii) the methods of acquisition, de la Cour Supérieure des Comptes et du Contentieux conservation, allocation and the accounting of Administratif, Article 2, 24 du 10 mars 2006. 72 BETTER SPENDING, BETTER SERVICES investment system under the assumption that Figure 5.12: Monitoring and Evaluation on Projects, ministries exercise appropriate ex-ante control in PIP, 2011-13 over the individual investment projects. As (in Number of Projects) previously noted, it also exercises an ex-ante role 900 in the validation of contracts.47 800 700 5.46. The requirement for a tripartite (SAI, MPCE 600 and line ministry) audit team to conduct 500 a closing audit of projects has never been 400 enforced. The regulatory framework requires 300 that a project be audited at the completion by 200 a team composed of representatives of the SAI, 100 0 the line Ministry and MPCE after a completion # projects inPIP Physically monitored by MPCE Audited by IGF Audited by CSCCA report is produced by the line ministry; however, there are no sanctions for non-compliance. The SAI has not yet performed financial closing audits at project completion nor did it request 2012 2013 any project completion reports. The SAI receives Source : Haitian authorities no information of the history of the project’s execution, no copies of the quarterly audits that are to be conducted by the responsible ministry’s M&E unit, its technical division or by MPCE. 5.47. There is no significant ex-post control over the Government’s investment budget (Figure 5.12). Given the absence of an internal audit function, continuing budget and staff restrictions on ministries that limit physical and financial quarterly inspections and the inability of MPCE to adequately monitor projects, it is unlikely to see in the future a significant increase in ex-post controls (Box 5.5). 47 There is considerable debate within Haiti and among donors over the appropriateness and the utility of this ex-ante role. However, it continues to date. 73 Box 5.5 : Project Completion The closing of projects is an important component of the investment project accountability loop. The Figure below shows the performance of line ministries during this phase. Project Completion/Evaluation in our Sample (Number of Projects) 18 16 14 12 10 8 6 4 2 0 Ministry closing Asset Joint closing inspection transferred audit Yes No Ongoing or Inapplicable Source : Haitian authorities From the 20 projects surveyed, eight had undergone a closing inspection by the line ministry, the rest had not (either because they were on-going, operating budget projects included in the investment budget or were not registered with the reporting ministry). The regulations48 required a joint closing audit by a three-person team, composed of representatives from the ministry, MPCE and the SAI. Some ministries reported having gone through a joint audit. However, it was not possible to obtain evidence of these audits. Several ministries seemed unaware of their responsibilities for ensuring that completed projects are audited and the assets transferred to the ministry’s accounts. 48 Decree of October 4, 1984 creating the Public Investment Fund within the Ministry of Planning; and Arrêté of September 17, 1985, stating the execution of the October 4, 1984 decree. 74 BETTER SPENDING, BETTER SERVICES Chapter 6: Increasing Sustainability and Improving Further Health Outcomes49 This chapter discusses recent trends in health indicators in Haiti, the structure of health provision, and the role of public spending. Donor assistance to the health sector has substantially increased over the past decade with improvements in health indicators, and narrowing gaps between income groups and regions. The role of the State remains, however, very small and access to health services remain limited. In light of its heavy reliance on donor assistance and given the falling level of external resources, the health sector faces a particular challenge in ensuring a sustainable financing model for the future. Furthermore, the composition of public spending may not be conducive to better services with increases in operating expenditure not keeping up with the expansion in public investment. Health services are unevenly distributed geographically and the number of daily visits is low, possibly because of medical personnel holding second jobs. Going forward, the authorities are thinking about making greater use of results-based financing. A. Although Low, Health Indicators with the average of countries of similar levels of economic development and much lower than Have Improved …49 the average of fragile states. Compared with 6.1. Health indicators in Haiti are poor. Maternal other countries in Latin America and Africa, mortality rate in Haiti is almost five times higher utilization of health services remains also than the regional LAC average of 85 and the lower. For example, Haiti has the lowest rate of under-five mortality rate is also four times higher deliveries by skilled birth attendants (37 percent) than the LAC regional average of 18 (Figures 6.1 in the region and compared to low-and middle- and 6.2). These outcomes are, however, in line income African countries. The immunization coverage rate is also much lower in Haiti (45 49 Prepared by Eleonora Cavagnero, Marion Cros, and percent) than in other Latin American and Andrew Sunil Rajkumar. African countries (Figure 6.3). The diarrhea 75 Figure 6.1 : Maternal Mortality, 2013 Figure 6.2 : Under-5 Child Mortality, 2013 (Per 100,000 live birth) (Per 1,000 live births) 600 100 500 90 80 400 70 60 300 50 40 200 30 100 20 10 0 0 Fragile states LIC Haiti Caribbean small states LAC Fragile states LIC Haiti Caribbean small states LAC Source : World Development Indicators Source : World Development Indicators treatment rate is one of the services for which skilled birth attendants increased by 76 percent, coverage in Haiti is comparable or better than in institutional deliveries increased by 125 percent, other Latin American and African countries. the treatment of diarrhea increased by 87 percent, and immunization coverage by 50 percent (DHS, 6.2. Maternal and child health indicators have 1994-95 and 2012). improved in Haiti, however, over the last two decades. Despite the devastating earthquake of 6.3. The gap in health outcomes between rich and 2010, maternal and child health seem to have poor households has also narrowed. Significant improved over the last two decades: the under- gains have been made in infant mortality for the five mortality rate (U5MR) declined from 145 poorest quintile, as well as the second and third deaths per 1,000 live births in 1990 to 73 deaths quintiles over the last decade (Figures 6.8 and per 1,000 births in 2013 (Figure 6.4). In addition, 6.9). Meanwhile, surprisingly, infant mortality the maternal mortality ratio (MMR) between increased slightly for the fourth quintile, as 1990 and 2013, from 670 deaths per 100,000 well as for the richest quintile in 2012 (DHS, live births in 1990 to 380 deaths per 100,000 live 2012). This could be the result of the earthquake births in 2013 (Figure 6.5).50 Haitians can now which affected the Metropolitan area where expect to live 8 more years than back in 1990, households are relatively richer compared to two additional years gained compared to the the rest of the country. Reduction in stunting average LAC country (Figure 6.6). All health and in the prevalence of diarrhea has been more service utilization indicators improved (Figure pronounced for poor income groups (Figures 6.7): Between 1994-5 and 2012, deliveries by 6.10 and 6.11). 50 MMR numbers, estimates from the World Health Organization, are however, not as reliable as those for IMR and U5MR, calculated from household surveys. 76 BETTER SPENDING, BETTER SERVICES Figure 6.3 : Health Service Utilization (In Percent of Population) 67 Burkina Faso 66 (2010) 48 81 59 Ghana (2008) 57 53 79 84 Benin (2012) 87 54 48 44 Kenya (2009) 43 72 68 83 Honduras 46 (2012) 60 85 93 Guyana 89 (2009) 59 56 37 Haiti (2012) 36 58 45 0 20 40 60 80 100 Skilled birth attendance Institutional delivery Diarrhea treatment Immunisation Sources : Demographic and Health Surveys and World Health Organization OMS 77 Figure 6.4 : Changes in Under-5 Child Mortality Rates, Figure 6.5 : Maternal Mortality Ratio, 1990-2013 1990-2013 (Per 100,000 live births) (per 1,000 live births) Caribbean small 700 Fragiles states 650 states Haiti 600 LAC LIC 0 550 -10 -20 500 -30 -40 450 -50 -60 400 -70 350 -80 -90 300 -100 1990 1995 2000 2005 2013 Source : World Development Indicators Source : World Health Organization (WHO) Figure 6.6 : Changes in Life Expectancy at Birth, Figure 6.7 : Health Service Utilization, 1994-2012 1990-2012 (In Percent of Population) (In Years) 9 30 8 Inmunization 33 41 7 45 6 5 31 4 Diarrhea treatment 41 44 3 58 2 1 16 0 Institutional delivery 17 22 36 Haiti LIC LAC Fragile states Caribbean small states 21 Skilled birth attendance 24 26 37 0 20 40 60 DHS 1994-95 DHS 2000 Source : World Development Indicators DHS 2005-6 DHS 2012 Sources : Demographic and Health Surveys and World Health Organization 78 BETTER SPENDING, BETTER SERVICES Figure 6.8 : Change in Infant Mortality, 2006-12 Figure 6.9 : Change in Under-5 Child Mortality, (Percentage Points) 2006-12 (Percentage Points) 20 20 15 15 10 10 5 5 0 0 -5 -5 -10 -15 -10 -20 -15 -25 -20 -30 Q1 Q2 Q3 Q4 Q5 Total Q1 Q2 Q3 Q4 Q5 Total Sources : Demographics and Health surveys Sources : Demographics and Health surveys Figure 6.10: Change in Stunting Rate, 2006-12 Figure 6.11: Change in Diarrhea Prevalence Rate, (Percentage Points) 2006-12 (Percentage Points) 0 0 -2 -1 -4 -2 -6 -3 -4 -8 -5 -10 -6 -12 -7 -14 -8 Q1 Q2 Q3 Q4 Q5 Total Q1 Q2 Q3 Q4 Q5 Total Sources : Demographics and Health surveys Sources : Demographics and Health surveys B. … But Disparities Remain … Respiratory Infections (ARIs) improved for the lowest quintiles and ended up matching those in 6.4. Income disparity in health outcomes remain. the highest quintiles (DHS, 2005-6 and 2012). Despite some improvements, the 2012 DHS The prevalence rate of ARIs was thus 14 percent generally shows large inequalities in health for children in the lowest quintile and 13 percent outcomes and results that are worse for the for children in the highest quintile in 2012, poorest quintiles. Infant mortality was thus whereas in 2005-6, the children in the lowest found to be 62 (per 1,000 live births) for the quintile had a prevalence rate of ARIs that was highest quintile of income, versus 104 for the twice as high as that of children in the highest lowest quintile of income in 2012. Compared quintile (DHS, 2005-6 and 2012). to the highest quintile, the number of children 6.5. The poorest benefit the least from health suffering from stunting was four times higher services, despite improvements over time. The in the lowest quintile in 2012 (DHS, 2012), a children’s vaccination coverage and diarrhea gap that had not changed from 2006. However, treatment rates seem to be more consistent the prevalence rates for diarrhea and Acute 79 Table 6.1 : Maternal and Children Health Coverage by Income Group, 2005-6 and 2012 (In Percent of Population) Q1 Q2 Q3 Q4 Q5 Total DHS 2005-6 Immunization 34 40 45 37 56 41 ARI treatment 27 31 41 40 40 35 Diarrhea treatment 34 38 47 54 54 44 Skilled Birth — — — — — 54 Skilled Attendant — — — — — 26 Delivery Health Facility 5 8 17 35 58 22 Stunted 41 37 34 18 8 29 DHS 2012 Q1 Q2 Q3 Q4 Q5 Total Immunization 43 46 52 42 41 45 ARI 23 32 36 52 52 38 Diarrhea 57 52 59 61 62 58 Skilled Birth — — — — — 67 Skilled Attendant — — — — — 37 Delivery Health Facility 9 20 38 51 76 36 Source: Demographic Health Services, 2006 and 2012 across wealth index quintiles in 2012 than in and an under-five mortality rate of 108 deaths 2005-6 (DHS, 2005-6 and 2012), but inequalities per 1,000 births, or 1.17 times the national persist for the other health services. 52 percent average (Figure 6.12). In terms of medical of children with ARIs received treatment in coverage indicators, The Metropolitan area the richest quintile against only 23 percent of has the highest institutional delivery rate and children with ARIs in the poorest quintile (DHS, the lowest underweight rate. Immunization 2012). In addition, coverage of institutional coverage remains similar to the national average deliveries was eight times higher for the richest (34 percent), while most departments are doing quintile (76 percent) than for the poorest better than Port-au-Prince in this area. The quintile (9 percent) in 2012 (DHS, 2012), which department of the Centre also has infant and highlights the fact that the poorest have very under-five mortality indicators that are more limited access to maternal health services. alarming than the national average. In contrast, the department of the North-West presents 6.6. Geographical disparities have not disappeared the best child health indicators, with an infant either. The Metropolitan area has the highest mortality rate of 41 deaths per 1,000 births and infant mortality rates, due to the earthquake, but an under-five mortality rate of 57 deaths per the service coverage is generally better compared 1,000 births (DHS, 2012). to other departments. The Metropolitan area has an infant mortality rate of 81 deaths per 1,000 births, or 1.26 times the national rate, 80 BETTER SPENDING, BETTER SERVICES 6.8. While the level of total health expenditure Figure 6.12: Child Health Outcome Indicators, 2012 per capita is relatively high in Haiti, public (Per 1,000 Live Births) spending in health is very low. In 2013-14, Haiti was allocating only 5 percent of government Nord-Ouest Nippes spending to health, well behind the 15 percent Grande-Anse recommended by the Abuja Declaration, and Sud government spending accounted for only about Centre Artibonite 1.5 percent of GDP, a low rate compared to other Nord-Est low- and middle-income countries (Figures 6.15 Nord and 6.16). Senegal, Benin, Bolivia, and Ghana – Sud-Est Ouest countries broadly at the same development level Metropole and facing similar health challenges – allocate Nationale average 10 percent of government expenditure to health, 0 20 40 60 80 100 120 and in the case of Rwanda, this figure reaches under-5 mortality rate 22 percent. Domestic expenditure on health infant mortality rate represents between 2 and 6 percent of GDP in Source : ICF International, 2012. The DHS Program other low- and middle-income countries (WHO, STATcompiler - http://www.statcompiler.com 2014). While the share of government spending allocated to health in Haiti was above the average in LAC and LIC countries, it plummeted in the aftermath of the earthquake. This event marked a C. … With Relatively High Health substantial change in the financing of the sector (Figures 6.17 and 6.18). Spending Financed by Donors 6.9. Out-of-pocket expenditure has been recently 6.7. Total health expenditure per capita is higher in rising substantially, as donor assistance has Haiti than in other low-income countries, but declined. Following the earthquake, donor still much lower than the average of countries assistance in the health sector has surged. in the region. Health expenditure per capita in Donors represented 28 percent of total health Haiti has increased in line with the average of expenditure in 2010. This share rose to 90 percent low-income countries and at USD76.6 stood in in 2011, following the earthquake. Since then, 2012 above what the average low-income country donor contributions have progressively declined, spends (Figures 6.13 and 6.14). According to the covering about 26 percent of health spending in WHO, USD50 would ensure the provision of 2013 (Figure 6.19), questioning the sustainability a basic care package to the population (Sachs, of the recently-achieved progress. This situation 2001).51 Haiti reached that target, spending puts more financial pressure on households with 1.5 times that amount in 2011-13. This level of out-of-pocket expenditure rising substantially. expenditure was, however, financed mainly by Households only paid 5 percent of health donors. spending in 2011. This share has expanded to almost 50 percent in 2013. This development is alarming in a country where about 60 percent of the population live below the poverty line. 51 WHO estimated US$34 in 2001 prices which would correspond to US$50 in 2014 prices, calculated with a similar methodology than Cavagnero et al. (2008). 81 Figure 6.13: Health Expenditure Per Capita, 2012 Figure 6.14: Health Expenditure Per Capita PPP, (Current US$) 1995-2012 (Constant 2005 US$) 1000 120 900 800 100 700 600 80 500 60 400 300 40 200 100 20 0 LAC Caribbean Fragile Haiti LIC 0 2011 1996 1997 1998 1999 2000 2001 2002 2003 2004 2006 2007 2008 2009 2010 2012 1995 2005 small states states Haiti LIC Source : World Development Indicators Source : World Development Indicators Figure 6.15: Public Health Expenditure, 2012 Figure 6.16: Public Health Spending, 2000-12 (Percent of GDP) (Percent of Government Expenditure) 4,5 18 4,0 16 3,5 14 3,0 12 2,5 10 2,0 8 1,5 6 1,0 4 0,5 2 0,0 0 LAC Caribbean Fragile LIC Haiti 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 small states states Haiti LAC LIC Source : World Development Indicators Source : World Development Indicators Figure 6.17: Health Expenditure, 2005-10, Figure 6.18: Health Expenditure, 2011-12 (Percentage of Total) (Percentage of Total) Public domestically Out- of- pocket Public domestically Out- of- pocket funded 39% funded 16% 26% 8% External External resources resources 35% 76% Source : World Development Indicators, World Health Organization Source : World Development Indicators, World Health Organization and MSPP 82 BETTER SPENDING, BETTER SERVICES from seeking medical attention in other low- Figure 6.19: Health Expenditure by Source, income countries (Figure 6.20). After Burundi, 2004-2013 (% of Total Health Expenditure) Haiti is the country where the lack of money is 100% most frequently cited as a barrier. 80% 6.11. In Haiti, risk-pooling mechanisms are weak. Only 4 percent of the population is covered by 60% a health insurance system in Haiti (World Bank 40% and ONPES, 2014). This rate is very similar to that of other low-income countries such as 20% Benin (Benin DHS, 2010) and Mali (Mali DHS, 0% 2012), where only 3 percent of the population is 2013 2011 2004 2006 2007 2008 2009 2010 2012 2005 covered by health insurance (Figure 6.21). Low- income countries such as Rwanda, on the other Out-of-pocket hand, show a very high enrollment rate, since External resources 92 percent of Rwanda’s population is enrolled Public domestically funded in a mutual health insurance program. Such a Source : World Bank, World Development Indicators system allows poor households to be covered without the burden of out-of-pocket spending at the point of care (MOH, 2010), thus protecting the poor against the financial risks caused by health problems. Donors fund, however, almost D. … Access to Health Is Limited half of the mutual health system in Rwanda Especially For the Poor … (Lagomarcino, 2012). Given the large informal sector in Haiti and the fragmentation in donor 6.10. Despite this recent increase in donor funding, traditional risk-pooling mechanisms assistance, lack of finance still prevents the poor may be difficult to implement. A more workable from seeking treatment. The surge in donor option could be to exempt specific groups or assistance seems to have been accompanied health services at public facilities: an option that by a substantial decline in the contribution of would require, however, public money. households to their health expenditure. This development should have improved access to 6.12. As a result, public facilities are not necessarily health services for the poorest. Yet, the most used by the poor. Dispensaries serve mainly recent household survey reveals that 65 percent rural areas, which are poor areas in majority. of households in the lowest consumption Yet, among households going to the public quintiles did not seek medical attention due to dispensary 49 percent are non-poor (Table 6.2). their lack of money against 39 percent for the Furthermore, hospitals serve a majority of noon- highest consumption quintiles (World Bank poor: 53 percent of households going to public and ONPES, 2014). Moreover, lack of money for hospitals are non-poor. This could be the result treatment is most frequently cited as a barrier to of higher payments required in public hospitals. health care in Haiti by women aged 15-49 (76 By contrast, community health interventions percent). The second problem faced by these seem to be pro-poor to a greater extent: 66 women is distance (43 percent) (DHS, 2012). percent of the beneficiaries of these services Distance and lack of money are also the most are poor (moderate and extreme poor). Finally, frequently cited barriers preventing women among households buying medications from 83 Figure 6.20: Barriers to Health Care Access for Women Aged 15-49, 2012 or 2013 (Percentage of Total) 90 77 76 80 72 70 62 60 52 52 47 44 44 50 40 41 43 36 34 37 40 29 27 30 25 23 24 19 21 18 21 20 11 8 9 10 2 0 Liberia Mali Benin Tanzania Burkina-Faso Burundi Haiti Obtain permission Lack of money distance Not willing to go alone Source: Haiti: MSPP, IHE and ICF International, 2013; Benin: Calverton, Maryland, USA: INSAE and ICF International, 2013; Mali: CPS, INSTAT, INFO-STAT and ICF International, 2014; Burkina Faso: INSD and ICF International, 2012; Liberia; Tanzania; NBS and ICF Macro. 2011; Burundi: ISTEEBU, MSPLS, and ICF International. 2012. Note: the total per country is not equal to 100 percent as respondents can give several answers. street vendors, the majority are poor (43 percent Figure 6.21: Enrollment of Populations in are even extremely poor), indicating that poor Risk-Pooling Systems, 2012 or 2013 (Percentage of Total) households may prefer self-treatment rather than seeking treatment at public facilities. 100 92 6.13. Catastrophic health expenditures (CHE) 80 in Haiti are more important among poor households. When people have to face out- 60 54 of-pocket spending, the amount can be so high in relation to income that it results in a 40 “financial catastrophe” for the household. These expenditures are not necessarily related to 20 12 10 catastrophic events (such as accidents or natural 4 3 3 disasters) but to the sizeable burden on household 0 welfare, as they may cut down on necessities Rwanda Ghana Honduras Burundi Haiti Benin Mali such as food and clothing, or are unable to pay for their children’s education (WHO, 2005). Sources : Lagomarcino (2012) and Demographic Estimating CHE by using a 25 percent threshold and Health Surveys of non-food expenses suggests that 3.4 percent of households incur CHE (Figure 6.22).52 This rate is two to three times higher among poor households than among non-poor households: 52 This threshold is in line with the consensus among the WB and the WHO (WB-WHO, 2013). These estimates were calculated using the 2012 ECVMAS. 84 BETTER SPENDING, BETTER SERVICES Table 6.2. Participation Incidence by Income Groups, 2013 (in Percent of Beneficiaries) Public Public CHW & Traditional Private Pharmacy Ambulant Other Dispensary Hospital Mobile Healer Providers Drug Clinics Seller Non-Poor 49 53 34 33 68 74 38 46 Moderate Poor 19 13 34 15 9 5 19 23 Extreme Poor 32 34 32 53 23 21 43 32 Total 100 100 100 100 100 100 100 100 Source : ECVMAS 2013 the CHE rate was 3.7 percent among moderately Figure 6.22: Incidence of Catastrophic Health poor households, 5 percent among the extreme Expenditures at Household Level, 2012 poor, and 1.7 percent among the non-poor. (Percentage) This rate was three times higher in rural areas 3.4 Men (5.0 percent) than in urban areas (1.6 percent), Women 3.5 suggesting that the poor and households in rural areas are more vulnerable to health shocks than Urban 1.6 the non-poor and households in urban areas. Rural 5.0 Lowest wealth quintile 4.1 E. … And Public Spending Is Not 6.1 2nd wealth quintile Geared to Better Services. 3.2 3rd wealth quintile 6.14. The composition of public spending may 4th wealth quintile 2.1 not be conducive to better service delivery. 1.5 Highest wealth quintile As noted in Chapter 3, overall operational Extreme poor 5.0 spending has not kept up with the recent rise in public investment. The health sector is a good Moderate poor 3.7 example. The investment budget from public Non-poor 1.7 funds (Treasury and Petrocaribe) in this sector National average 3.4 has doubled, going from 0.07 percent of GDP to 0.15 percent of GDP between 2006 and 2013. By 0 1 2 3 4 5 6 7 contrast, the increase in operating expenditure Source : World Bank estimates based on ECVMAS, 2012 and ADePT. has been more modest, amounting to 0.52 percent of GDP in 2006 and 0.7 percent in 2012. 6.15. Moreover, the wage bill may be crowding out critical operating expenses. From 2006 to 2012, the share of the operating budget allocated to personnel costs represented about 90 percent (Figure 6.23). In order to improve health outcomes, however, not only personnel is needed, but also and more importantly the 85 right mix of personnel and other inputs such 58, 53, 43 and 22 percent in Tanzania, Uganda, as medicine and equipment. In comparison, Burkina Faso and Benin). Non-salary operating other low- and middle-income countries expenditures include spending such as vaccines allocate a smaller share of their operating and medicines, i.e. critical elements of the health budget (excluding investment budget from system. Yet the share of expenditure allocated to public funds) to human resources (65 percent in consumer goods declined from 6 percent in 2006 Honduras, 58 percent in Ghana and respectively to 4 percent in 2012 (Box 6.1).53 Box 6.1 : Health Care Provision in Haiti Health in Haiti is managed by a three-tier system: • The primary health sector is managed by the District Health Unit (UAS). The UAS corresponds to the administrative division of the country and oversees the primary level: Community Referral Hospitals (HCRs), CALs, CSLs and dispensaries. Dispensaries, CSLs and CALs provide primary health services related to maternal and child health, family planning, medical emergency and the treatment of communicable diseases (TB, HIV/AIDS, Malaria), as well as prevention and health promotion services. In addition, each of these three types of institutions provide community health services through assembly stations, mobile clinics and home visits conducted by community agents, nurses/birth attendants, and matrones (traditional birth attendants). HCRs offer four types of health services in the following specialties: internal medicine, pediatrics, gynecology and obstetrics, and surgery. • The secondary level is managed by the departmental health directorate (DDS), which provides technical support to departmental hospitals, which are the referral hospitals at the departmental level. In principle, patients who cannot be treated by HCRs are referred to departmental hospitals. These offer more specialized services than HCRs, such as ophthalmology, urology, dermatology and orthopedics. • Finally, the tertiary level corresponds to the central level of the MSPP, which plays a regulatory role, is in charge of developing health policies and strategies as well as standards and procedures, and plays a monitoring and evaluation role. It directly oversees academic institutions or hospitals as well as the other two levels. The MSPP is also required to supervise the actions of the private for-profit sector even if in practice the MSPP has little information on the private sector. 53 4 percent of donor commitments were assigned to projects supplying inputs and medicines, which remain very low. Programs to combat endemic diseases (11 percent of the national investment budget) and priority health programs (9 percent) also include activities for the procurement of medicines and laboratory inputs, increasing the share of the commitments of national investment programs going to medicines or non-salary operating expenditures. 86 BETTER SPENDING, BETTER SERVICES Governance of the Haitian Health System Level Administration Public Private Ministry of Public Instituts Private Tertiary Health and Universitaires hospital Population et Spécialisés Departemental Hôpitaux Private Secondary Health Directorate Départamentale hospital (DDS) (HD) 2nd level: Community Referral Health Unit District Hospital (HCR) Private Primary (Unité dispensaries, level d’Arrondissement 1er Level: Health private de la Santé - UAS) Center w/o beds (CSL) clinics and Health Center w/beds (CAL) 1st Level: Dispensary Source : World Bank, adapted from the 2012 Health Master Plan The first-level primary sector accounts for three-quarters of the provision of care in Haiti. There were 907 health facilities in 2013 in Haiti (IHE and ICF International, 2013), 359 dispensaries, 298 CSLs and 129 CALs, which represents 87 percent of the total provision of care. 11 percent of the 907 health facilities were HCRs (second-level primary sector), and the remaining 2 percent were departmental hospitals (8) and University Hospitals/Institutes (8), mainly in Port-au-Prince. The for-profit sector plays a substantial role in the provision of primary health care services. While the for-profit private sector manages 24 percent of the overall provision of care, it focuses on primary care services. The for-profit private sector provides 29 percent of CSLs and 31 percent of HCRs, while there is no for-profit departmental hospital and only 1 in 8 hospital/ university institutes is a private for-profit one. The public sector includes the health facilities that are managed by the MSPP, the non-profit sector (structures managed by NGOs and the Church) and the mixed sector (managed by the MSPP and NGOs). Facilities run by the MSPP account for 38 percent of the provision of health care, those managed by the non-profit sector account for 18 percent, and mixed-sector facilities account for 20 percent (IHE and ICF International, 2013). 87 Box 6.2 : Definitions of Operating Budget Terms Personnel costs: sums spent on basic salaries and wages before deduction of tax withholdings at source, and of contributions to social insurance and civil pension funds, which are at the expense of employees. Compensation payments, premiums and bonuses are planned. Service expenditures and miscellaneous charges: these notably include basic services fees (communication, provision of water, electricity, and gas), promotion costs (printing costs), transportation and travel expenses, training costs, provision of services by third parties, intermediary fees and expenses, petty cash service expenditures to be broken down. Consumer goods and small equipment: supplies and small equipment including office equipment and surgical, medical, and pharmaceutical tools; chemicals and energy supplies, subsistence products, and textiles and apparels. Tangible assets: this is capital expenditure for payments made for the acquisition of movable fixed assets (furniture, machinery and equipment) or non-depreciable property (land) or depreciable property (buildings, edifices, etc.). Grants, assessments, contributions, benefits, compensation: a) grants (non-repayable unrequited payments that the State performs for current purposes or to achieve policy objectives or various objectives; b) assessments and contributions: expenses related to the share that the State has committed to pay for membership in various organizations or institutions; c) benefits: unrequited sums allocated by the State for social assistance benefits, incentives and aid, scholarships and awards, and grants of any kind; d) compensation: exceptional and unforeseen expenses. Other public spending: in this category are grouped ordinary expenses whose nature does not allow their classification in the above categories. Source: Government of the Republic of Haiti. Nomenclature of State budget expenditures and User guide. 2001 6.16. Furthermore, public investment may not be given the current fiscal space constraints of the targeted where it is most needed. In 2012, for health sector, these costs may not be covered. instance, nearly a third of the commitments of 6.17. Health services seem also to be unevenly national investment funds were allocated to the distributed geographically. The results of the construction or rehabilitation of hospitals (Figure most recent household survey would suggest 6.24). Only 22 percent of the commitments were that metropolitan and urban populations are allocated to investment projects for dispensaries, the main beneficiaries of public expenditure on health centers (Centre de santé sans lit, CSL and health. According to this survey, 51 percent of Centre de santé avec lit, CAL), although these the metropolitan population and 45 percent of facilities account for 87 percent of health care the urban population consult a public provider delivery in Haiti (Box 6.2). Focusing national (funded by the government or donors) against investment programs on hospital construction 43 percent of the rural population, which may not be sustainable in the long run. Indeed, comprises a majority of the country’s poor the construction of new facilities will require (Figure 6.25). However, the metropolitan and additional needs in operating costs (e.g. urban populations (respectively 36 percent and personnel, maintenance) over the long term, and 33 percent) tend to consult private providers 88 BETTER SPENDING, BETTER SERVICES Figure 6.23: Salaries, Excl. Domestic Investment Figure 6.24: Investment Budget from Public Funds, Budget 2012-13 (Percent of Government Budget) (Percentage of Total) 100 Programs against Priority health 90 endemic 9% illnesses Equipment 80 procurement 11% 70 at health facility level 60 3% 50 CSL (Construction) 40 10% Drugs and medical supply 30 4% 20 CAL (development) Governance 10 12% 0 19% Haiti (2012) Honduras (2010) Tanzania (2012) Ghana (2010) Uganda (2006-7) Burkina Faso (2005) Benin (2004) University HCR (development) hospitals and 25% specialized centers (Contruction) 7% Source : World Bank Sources : BOOST and World Bank Sta calculations Figure 6.25: Utilization of the Di erent Types of Structures by Area, 2013 (Percentage of Total) 100% other 90% Traditional birth attendant 80% 70% Ambultory drug seller 60% Mobile clinic 50% Pharmacy and lab 40% 30% Traditional healer 20% Community health workers 10% 0% Private providers Méeropolitan urban rural Public providers area Source : World Bank calculations based on ECVMAS, 2013. 89 more often than the rural population (23 times lower in rural areas than in urban areas. percent). Moreover, households in rural areas However, the density of health agents is 1.3 have a greater tendency to self-medicate and times higher in rural areas than in urban areas. consult traditional healers. There are 11.4 doctors, nurses and midwives per 10,000 inhabitants in urban areas against 6.18. Both the density of medical staff as well as 4.3 in rural areas (Figure 6.30). The density of hospital beds is low relative to low-income paramedical personnel is also higher in urban countries and the regional average (Figures areas than in rural areas. The density of health 6.26 and 6.27). There are 9.5 medical staff per agents is 4.1 per 10,000 inhabitants in rural areas 10,000 inhabitants in Haiti (counting medical and 3.1 per 10,000 inhabitants in urban areas, staff in private-for profit facilities), in line with which is explained by the fact that rural areas the average of Sub-Sahara Africa countries, but can be geographically more difficult to reach. much lower than the regional average and also Inaccessible zones are indeed more numerous in much lower than the average of countries at the rural areas, requiring a higher number of health same stage of economic development. Similarly, agents. the density of hospital beds is low. Haiti has 7 beds per 10,000 inhabitants below the number 6.21. The productivity of the medical personnel observed on average in the region and in of first-level primary facilities is low. A economy of a similar level of development. recent study was conducted on technical efficiency for a sample of 45 first-level primary 6.19. There are also disparities in the departmental facilities (dispensaries, CSLs and CALs). The distribution of medical personnel and beds. sample consisted of 11 CALs, 21 CSLs, and The departments of the South-East, Artibonite, 13 dispensaries in the departments of the West, Grand’Anse, North, North-West and Nippes the North-East, the North-West and the Centre have a density of doctors, nurses, and birth for 2011 (Box 6.3).54 Institutions that have a attendants, which is below the national average technical efficiency score of one are efficient. (Figures 6.28 (a) and (b)). The department of the The study shows that only 13 percent of health West has the highest density of doctors, nurses facilities (6 out of 45) produced health services and midwives per 10,000 inhabitants, followed efficiently: they had neither excessive operating by the North-East (9 per 10,000 inhabitants), costs, nor excessive personnel. The rest of the the Centre (8.9 per 10,000 inhabitants), and the facilities (with a technical efficiency score of South (8.8 per 10,000 inhabitants). In addition, less than one had an excess of either personnel the departments of the North-East, South-East, or operating costs for the number of services Artibonite, and Grand’Anse, have a low number provided (Figure 6.31). The average score for of hospital beds for a population of 10,000 Haiti was 0.48 low compared to similar studies inhabitants, with respectively 2.7, 3.3, 3.6 and on the technical efficiency of the primary health 3.7 hospital beds per 10,000 inhabitants (Figure sector carried out in other low-income countries 6.29). In contrast, the departments of the South, (Table 6.1). Furthermore, health facilities in West and Centre have a higher number of beds the metropolitan area tended to have a higher than the national average for public institutions technical efficiency score than facilities in rural (which is 5.6 beds per 10,000 inhabitants). and urban areas because they have more patients. 6.20. There are disparities in the distribution of medical and paramedical personnel between 54 This study on technical efficiency was based on the urban and rural areas. The densities of medical number of visits (output), and total expenditure and and paramedical personnel are 2.5 and 2 total number of personnel (input). 90 BETTER SPENDING, BETTER SERVICES Figure 6.26: Density of Medical Personnel, 2013 Figure 6.27: Density of Inpatient Beds, 2013 (Per 10,000 Inhabitants) (Per 10,000 Inhabitants) 100 71.5 30 90 80 24 25 70 21 60 20 50 40 15 30 14.9 20 6.9 9.1 10 7 10 20.4 0 2.6 5.1 2.5 5 Haiti Low Income Latin America Sub-Saharan Countries and the Africa 0 Caribbean Haiti Low Income Latin America Medical Dcotor Nurses Contries and the Caribbean Source : WHO-Health Statistics, 2013 Source : WB estimates (Haiti) based on SPA dataset (2013) and WHO 2013 for other countries. Figure 6.28: Density of Medical Personnel by Depart- ment – Private for-Pro t Sector Excluded a. Density of medical personnel, 2013 b. Density of pharmacists and community health agents, 2013 (per 10,000 inhabitants) (Per 10,000 Inhabitants) 8 10 0.4 0.1 7.5 0.4 0.1 7 9 0.1 0.2 8 0.1 6 2.6 5.2 5.2 7 0.1 0.1 4.7 3.6 5 0.1 0.1 3.6 4.4 6 2.3 2.7 2.1 4.1 4 3.6 5 4.3 2.5 3.2 3.0 2.9 1.4 1.6 2.1 4 3.3 3 2.9 1.1 1.6 2.4 2.2 2.2 3 2.7 3.4 3.1 2 3.3 3.6 1.4 2.2 0.9 1.8 2 1 1 0.3 0.1 0.2 0.1 0.1 0.2 0.2 0.2 0.1 0.2 0.2 2.9 1.6 1.8 2.1 1.5 1.5 2.0 0 0 Ouest Sud-Est Nord Nord-Est Artibonit Centre Sud Grand'Anse Nord-Ouest Nippes National West South-East North North-East Artibonite Centre South Grand'Anse North-West Nippes National Pharmaciens Agents de santé communautaires Doctor Nurse Auxiliaire-Nurse Midwive Source : calculs de la Banque mondiale Source : World Bank calculations based on the SPA dataset basés sur série SPA (IHE et ICF International, 2013). (IHE and ICF International, 2013). 91 Figure 6.29: Number of Beds, exc. Private For-Pro t (Per 10,000 Inhabitants) 9 8.5 8 7.1 7 6.1 6 5.5 5.6 4.6 5 4.0 3.3 3.6 3.7 4 2.7 3 2 1 0 Nord-East Sud-East Artibonite Grand-Anse Nippes North North-West South West Centre National Source : World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). Figure 6.30: Density of (a) Medical (b) Paramedical Personnel, and (c) Community Agents by Location, Excluding the Private For-Pro t Sector, 2013 (per 10,000 Inhabitants) Density of Medical Personnel Density of paramedical personnel 12 0.30 10 0.25 8 0.20 6 0.15 4 0.10 2 0.05 0 0.00 Urban Rural National average Urban Rural National average Medecal doctor Nurse Auxiliaire-Nurse Midwife Pharmacist Dentist Source : World Bank calculations based on the SPA dataset Source : World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). (IHE and ICF International, 2013). Density of community agents 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Urban Rural National average Source : World Bank calculations based on the SPA dataset (IHE and ICF International, 2013). 92 BETTER SPENDING, BETTER SERVICES Table 6.3 : Technical Efficiency In Haiti And Other Low-Income Countries Country Percent of Sample That Is Non- Average Score Sample Author Efficient (<1) Haiti 83 % 0.48 Random selection of 45 health facilities Cros, M., Zeng, W. but 60 % have a score < 0.5 in 4 departments 2014 Burkina Faso - 0.86 25 primary health facilities Marshall et al., 2011 Ethiopia 75 % 0.57 60 health posts in 7 rural districts out Sebastian et al., of 35 2007 Ghana 78 % 0.88 Random selection of 86 health facilities Akzali. et al., 2011 Guatemala 71 % but 53 % 0.78 34 health posts Hernandez, 2013 have a score > 0.9 Source: Cros, M., and Zeng, W. (2014). consultations per medical personnel in Haiti is Figure 6.31: Technical E ciency Score, 1.5 times lower than the average in Liberia and 45 First-Level Primary Facilities, (0=lowest, 1=highest) approximately 2.5 times lower than in Cambodia and Rwanda (Figures 6.32 and 6.33). If a doctor 30 27 works 7 hours daily, or 420 minutes, this means Number of health facilities 25 that this doctor sees one patient per hour (420/6 20 = 60). It should be noted that nurses were 15 included, given that in dispensaries, there are 10 9 no doctors, and patients also consult nurses for 6 prenatal care and family planning in CSLs and 5 3 CALs. 0 <0.50 0.50 - 0.75 0.75-0.99 1 6.23. A second occupation could explain low Source : Cros M., Zeng, W. (2014). productivity. The survey on human resources conducted by the World Bank, USAID and the MSPP revealed that 30 percent of the medical personnel in the North-East, 49 percent in the Centre, and 49 percent in the North-West The average technical efficiency score was 0.36 in derived an income from sales activities or from urban areas, 0.42 in rural areas, and 0.82 in the a private practice. This was confirmed by focus metropolitan area (Cros and Zeng, 2014). groups with medical personnel and validated 6.22. In particular, the number of consultations by health department directors at a workshop is low. The results indicate that the national in Port-au-Prince in June 2013 (World Bank, average is six daily consultations per medical USAID, MSPP, 2013). It is thus possible that in personnel.55 By comparison, the number of daily some cases the medical personnel actually work part-time instead of full-time at facility level, but 55 A study by the World Bank, USAID, and the MSPP, are paid full time by an NGO or the MSPP, and confirms these results: in 2012 the medical personnel engage in other activities such as having their and health agents were carrying out an average of own practice, selling medicines or phone cards. four visits per day (institutional and community) in Such side activities increase their salary by at dispensaries and 6 visits per day in CSLs/CALs. 93 Figure 6.32: Consultations Per Doctor and Nurse, exc. Figure 6.33: Consultations per Medical Sta Private For-Pro t Sector, 2013 (Daily Numbers) (Daily Numbers) 12 25 10 10 10 20 8 7 6 6 6 6 6 15 6 5 5 4 10 4 2 5 0 0 Haiti Liberia Cambodia Rwanda Morocco West South-East North North-East Artibonite Centre South Grand'Anse North-West Nippes National (2013) (2009) (2008) (2011) (2014) Sources : SPA, 2013 (Haiti); USAID, 2009 (Liberia); Source : World Bank calculations based on the SPA dataset Royal Government of Cambodia, 2008 (Cambodia); (IHE and ICF International, 2013) MSH, 2011 (Rwanda); Kingdom of Morocco, 2014 (Morocco). Box 6.3 : Definition of Technical Efficiency Applied to the health sector, technical efficiency consists in achieving a maximum level of consultations or hospitalizations in a health facility with a given level of inputs (Street, 2011). There are 2 cases: 1) input-oriented technical efficiency, which aims at determining the proportion by which inputs (personnel, other expenses) must be used to produce a given number of consultations; 2) results-oriented technical efficiency, which studies the additional number of consultations possible without having to change the health facility’s number of inputs (Coelli, 1996). The use of linear programming, known as data envelopment analysis (DEA), is a non- parametric method that determines the number of healthcare facilities included on an efficiency frontier. This method gives a technical efficiency score based on the number of inputs, such as personnel, current expenditures, and results, i.e. consultations. The technical efficiency score is comprised between 0 and 1. A score of 1 means that the health facility is on the efficient frontier and is efficient. A score below 1 demonstrates poor performance, especially if the score is close to 0. Initially applied in the industrial sector, this methodology is increasingly used in the health sector to measure the technical efficiency of hospitals or primary health care facilities. 94 BETTER SPENDING, BETTER SERVICES least 5 percent and in some cases double it. This Figure 6.34: International Comparisons-Share of might explain the low number of consultations Administrative Personnel and poor productivity of the medical personnel. In Total Number of Primary Health Facilities 6.24. Excessive administrative personnel may also (Percentage Total) 50 undermine efficiency. The same study shows 45 40 that administrative personnel represent 40 40 35 33 percent of total institutional personnel. In MSPP 30 29 30 facilities, administrative personnel represent 25 20 45 percent of institutional personnel and 39 20 15 percent of institutional personnel in non-profit 10 institutions (NGOs). There is no standard on 5 0 the number of administrative personnel, but in Haiti Afghanistan Rwanda Liberia Liberia (dispensaire) (CSL/CAL) general, the need for such personnel in primary level facilities is low. These facilities are small and Source : Haiti: Cros, M. and Zeng, W., 2014; Afghanistan : Transtional Islamic Government of Afghanistan, have little or no equipment (for dispensaries), 2003; Rwanda: USAID, 2011; Liberia: USAID, 2009 requiring a low level of maintenance and management, and therefore few administrative personnel. Furthermore, given the low daily productivity previously noted (6 visits per doctor the expansion of the RBF national strategy. The and nurse), the high proportion of administrative fact that health facilities have limited technical personnel is not justified. This ratio seems also efficiency can certainly be explained by their high in comparison to economies at similar very low productivity, having few patients. To development levels and facing similar health improve the use of their resources, it is therefore challenges (Figure 6.34) in the interest of such facilities to operate mobile 6.25. Going forward, the government aims at clinics to increase the number of consultations. implementing results-based financing (RBF) to RBF could provide monetary and non-monetary improve productivity and technical efficiency incentives to health workers in order to increase of primary-level facilities. Currently in a pilot their productivity, discouraging them from phase, the MSPP is working with partners on exercising another profession (Boxes 6.4 and 6.5). 95 Box 6.4 : More Health for Every Dollar: Results-Based Financing What is Results-based financing? Results-Based Financing (RBF) is an instrument that links financing to pre-determined results, with payment made only upon verification that the agreed-upon results have actually been delivered. RBF for health refers to any program that transfers money or goods to either patients when they take health- related actions (such as having their children immunized) or to healthcare providers, when they achieve performance targets (such as immunizing a certain percentage of children in a given area). What Makes Results-based Financing different? Traditionally funding for health has been directed toward inputs—salaries, construction, training, equipment. Improved health was assumed to follow, but this has not always happened. Despite billions of dollars over the last decade, many countries in Africa are still falling short, particularly in areas that require a functioning health system. Sub-Saharan Africa, for instance, has the highest rate of maternal deaths in the world with an average of about 900 deaths per 100,000 live births. Child deaths and malnutrition are also serious problems. The fundamental issue is the poor performance of the public health care system, including low levels of physical access in some places; poor quality of care; a lack of adequate incentive structures for health workers; weak management; and inadequate data of sufficient quality to monitor and evaluate progress. Individuals must demand services; health workers must be motivated to deliver adequate care; and the institutions they work for must be encouraged to make the systemic changes required to achieve health goals. RBF flips the whole equation on its head, starting with the result— more children immunized, for example—and letting health workers and managers on the ground decide how to achieve them. The Potential of RBF A number of developing country experiences strongly suggest that RBF can work. There are currently three countries (Rwanda, Burundi and Sierra Leone) with nationwide programs and 14 countries with ongoing pilots. These programs help improve health; strengthen health systems; spur innovation, creativity and country ownership; and encourage reforms that confer authority and flexibility to local service-delivery levels, fostering problem-solving where it is most needed. When poor patients or households have been offered financial or material rewards for adopting health-promoting practices, they respond and health indicators improve. Similarly, when health workers and facilities are given bonuses upon achieving targets, those targets tend to be met. Results- based financing has also been shown to help to increase patient demand for health services. In addition to improving health, results-based financing can also contribute to strengthening a country’s health information system. Because accurate monitoring and evaluation of RBF schemes require the development of robust health information and management systems, incorporating the RBF concept, even into donor funds aimed at specific diseases, reinforces efforts to improve the timeliness, credibility and accuracy of national reporting and monitoring, thus contributing to improving the overall capacity of a country’s health system. Source: L. Morgan (2012) 96 BETTER SPENDING, BETTER SERVICES Box 6.5 : The Right Incentives Lead to Measurable Results in Rwanda In an effort to improve maternal and child health, Rwanda began paying for performance at the health facility level in 2006. At the time, health workers and facilities were in short supply (only 36 hospitals and 369 health centers in a country of nine million people, and only one doctor per 50,000 inhabitants). Many people lacked access to care, and the quality of care was often low. In 2001, three non-governmental organizations working in Rwanda attempted to address the problem by raising health workers’ salaries. Nothing changed. Then they tried linking bonuses directly to performance – for example, if the health worker or facility could show that ten more women had given birth in a facility rather than at home where women risk dying from complication, they would receive a bonus. Paying for performance worked. Following three successful RBF pilots, the government of Rwanda designed and implemented a nationwide RBF scheme, folding a rigorous impact evaluation into the roll out. Results released in 2009 revealed significant improvements in the deliveries and preventive care visits by young children. Source: F. Niyuhire (2010) 97 98 BETTER SPENDING, BETTER SERVICES Chapter 7: Keeping children in school and improving education outcomes56 This chapter discusses recent trends in education indicators in Haiti, the structure of education provision, and the role of public spending. With greater donor assistance and an increasing role of the State through tuition waivers, education access has improved, but education indicators remain low with late primary school entry, high dropout, and limited learning. Households still spend a substantial amount for sending their children to school. Geographic disparities in service delivery remain with the West Region having more schools per capita, and quality is poor. Monitoring tuition waivers needs to be improved, and linked to better learning conditions in schools and actual learning in the classroom. Results-based financing offer an opportunity to maintain access, increase accountability, and improve learning, but better Monitoring and Evaluation systems are needed for this. A. Despite Progress, Education of the State in financing primary education in recent years through tuition waivers and school Outcomes are Low …56 canteen programs. 7.1. Educational attainment has been rising steadily 7.2. Today, the majority of children are in school. across cohorts of young adults. In 1994, fewer The majority of preschool age children, and 90 than 30 percent of 15-19 year olds had reached percent of children of official primary school age lower secondary school, a figure which had (6-11), are in school (Figure 7.1). In 2001, the risen to over 50 percent for women and over national net enrollment ratio for primary school 40 percent for men by 2012 (Measure DHS). stood at about 60 percent, and by 2012 had risen These gains are due in part to increased access to to 72 percent.57 Similarly, the overall secondary schools, as the number of low-cost, non-public schools has expanded, and to the increasing role 57 The discrepancy between the primary net enrollment ratio and the overall 90 percent enrollment rate of 6-14 year olds is due to the large number of primary 56 Prepared by Melissa Adelman. school age children in preschool. 99 Figure 7.1 : Share of 3-25 Year Olds Currently Figure 7.2 : Haiti’s Literacy Rate is between LIC and in School, 2012 LAC Averages (Percent of Total) 100 Taux d’alphabétisation (adultes âgés de 15 ans et plus) par département 90 80 - 85 80 75 - 80 70 70 - 75 65 - 70 60 60 - 65 50 55 - 60 40 91.5 % : LAC average 30 61.2 % : Low-income country average 20 77.5 % : Haiti average 10 0 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Sources : World Bank sta estimates using ECVMAS 2012 Notes: Data on literacy for countries other than Haiti comes from net enrollment ratio rose from 22 percent to di erent sources, and therefore methodological di erences may 47 percent. These increases reflect progress in a ect the comparisons between Haiti and other countries. Sources: World Development Indicators; World Bank sta both the share of children in school and the estimates using ECVMAS 2012 accordance between age and grade. 7.3. Despite this progress, educational attainment and learning are still relatively low. Compared to its LAC neighbors, Haiti has the highest share 15, but 73 percent of 18 year olds report still being of adults with no education, and the highest in school.58 In addition, substantial distortions share of 15-19 year olds who have not completed between age and grade remain, driving large primary school. Literacy rates in all departments, differences between net and gross enrollment including the West, are lower than the LAC rations at every level, until participation drops average, and in several departments are close to off steeply at the tertiary level Figure 7.3 (a) (b). the global average for Low-Income Countries (LIC) (Figure 7.2). Nationally, the adult literacy 7.5. Children from poor, rural households, living rate is about 77 percent, midway between the in departments far from Port-au-Prince, are LIC and the LAC averages. more likely to be over-age or out of school. To understand better the correlates of school 7.4. Haiti’s school enrollment still lags behind the enrollment and over-age status, probit regressions region and most children are over age for their were estimated (results are presented in Annexes grade. While the majority of children are now 2 and 3). Holding other characteristics equal, in school in Haiti, within the Latin America and for a HTG 1,000 increase in annual household Caribbean region, only Nicaragua (88 percent), per capita consumption (worth about 4 percent Guatemala (92 percent), and Honduras (94 of the national poverty line), the probability of percent) also have enrollment rates below 95 percent for this age group (SEDLAC 2014). 58 Enrollment here is based on answers to the survey School enrollment begins to drop off around age question asking if children are ‘currently in school’ rather than on administrative records. 100 BETTER SPENDING, BETTER SERVICES Figure 7.3 : Figures (a) and (b) - Gross and Net Enrollment Ratios, 2001-12 (Percent of Children in Primary School) a) Primary Net Enrollment Ratios, 2001-12 (Percent) b) Net and Gross Enrollment Ratios, 2012 (Percent) 100 140 90 80 120 70 60 100 50 40 80 30 20 60 10 0 40 East Asia & Paci c Europe & Central Asia Latin America & Caribbean Middle East & North Africa North America South Asia Sub-Saharan Africa Haiti 20 0 Primary Lower Upper Tertiary Secondary Secondary 2001 2012 NAR GAR Sources : UNESCO, Institut for Statistics; Sources : UNESCO, Institut for Statistics; World Bank EdStats; ECVMAS (2012) World Bank EdStats; ECVMAS (2012) school enrollment rises by 0.2 percentage points. school enrollment and household head education Department of residence also has a significant and household poverty (discussed above) effect on school enrollment: children living in demonstrate how poverty can persist from one Artibonite, West, and South-East (departments generation to the next through education. surrounding Port-au-Prince, and also with the 7.7. The primary level of the education system lowest poverty rates) are all more likely to be in is highly inefficient: children start primary school and less likely to be over age than children school two years late on average and fewer than living in other departments. Children from 60 percent will reach the last grade of the cycle. urban households are 2 percentage points more While the official age for beginning primary likely to be in school and 19 percentage points school is 6, the average child enters first grade for less likely to be over age than children living in the first time at 7.8 years old, after having spent rural areas. two or more years in some form of preschool 7.6. Children living in households with less (Table 7.1). This distortion grows over time, educated heads are also less likely to be in as about 10 percent of children repeat and 2-6 school and more likely to be over age. The percent drop out of each grade of primary. These results of the probit regressions also show repetition rates are higher than the averages in that children living in households where the the rest of LAC and Sub-Saharan Africa. Using a household head completed primary school are simulated cohort approach, these rates imply that 4 percentage points more likely to be in school only about 58 percent of children in first grade than children in households where the head had will arrive at sixth grade, and only 29 percent no formal schooling. They are also 23 percentage will reach the final year of upper secondary.59 For points less likely to be over age than children in households where the head had no formal 59 The simulated cohort approach used to calculate the schooling. These correlations between children’s survival rates is based on the UNESCO Institute for 101 Table 7.1 : Late Starts, Repetition, and Drop out Contribute to Low System Efficiency Grade Average Prescribed Age Percent Expected Percent Expected To Age To Repeat Drop Out 1 8.1 6 12 % 2% 2 9.9 7 10 % 1% 3 11.5 8 11 % 2% Primary 4 12.8 9 9% 3% 5 13.8 10 7% 3% 6 15.3 11 11 % 5% 7 15.9 12 6% 3% Lower 8 16.8 13 3% 4% secondary 9 17.8 14 10 % 5% 3 18.5 15 4% 4% Upper 2 19.6 16 4% 7% secondary Rétho 20.6 17 29 % 13 % Philo 20.8 18 9% 30 % Global comparisons average primary Sub-Saharan Africa 8.6 % repeaters Latin America & Caribbean 5.7 % Note: Percentages expected to repeat and to drop out after grade X are based on the share of individuals who were in grade X last year who are in grade X again this year (repeaters) and who are not in school this year (dropouts). Source: UNESCO Institute for Statistics; World Bank staff estimates using Demographic and Health Survey 2012. those who do eventually complete primary, the years old for boys. It may also be related to many average ages by grade suggest that it takes 7 to 8 other social and economic factors, including for years on average to complete 6 years of primary example childbearing and migration. school. Therefore, identifying and addressing 7.9. Furthermore, many students, particularly in the drivers behind late primary school starts, as poor communities, learn little. Assessments well as the high repetition and dropout rates, is administered in early grades in selected schools critical to increasing educational attainment. have found that fundamental skills are acquired 7.8. After age 14, girls start dropping out of school very slowly or not at all, particularly in schools slightly faster than boys. As Figure 7.4 shows, in poor communities. For example, assessments girls and boys are enrolled in school at roughly conducted in schools in the Artibonite and equal rates through age 14, and then girls begin Nippes found that the average third grader could to leave school more quickly than boys. This only read 23 words per minute, well below the trend is similar across individuals in poor and estimated speed of 35-60 words per minute non-poor households, as well as urban and required for comprehension of a basic text.60 rural areas. It may be due in part to girls’ faster Students sitting for the national exams are a progression: in the final year of primary school, relatively select group, considering that many the average girl is 14.9 years old compared to 15.6 children do not reach beyond sixth grade, and that sitting for the first two exams requires Statistics method: survival rate from grade A to B = paying a fee (250 gourdes in sixth grade; 350 (students in grade B – repeaters in grade B)/students in grade A. 60 Research Triangle Institute 2010; USAID 2012. 102 BETTER SPENDING, BETTER SERVICES in ninth). 2013 passing rates were about 75 Figure 7.4 : Youth Enrolled in School percent at grades 6 and 9, falling to 29 percent in by Age and Gender, 2012 Rheto and 38 percent in Philo. These rates vary (Percentage of Total) by department, as does the share of students 100 90 actually sitting for the exams (Table 7.2). Given 80 the weaknesses in basic skills suggested by 70 60 small-scale studies, nationally representative 50 learning assessments at earlier ages are needed to 40 30 understand the challenges faced by the majority 20 of Haitian students. 10 0 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Female Male B. … Driven by a Greater Sources : World Bank sta estimates using ECVMAS 2012. Education Supply from the Non-Public Sector … 7.10. Growth in education supply has primarily come from the non-public sector. While all children in Haiti are constitutionally entitled to a basic education, the State has historically not fulfilled this obligation. Formal education in Haiti is structured in four levels and, in practice, the majority of education services at all levels have been provided by the non-public sector (Box 7.1). Figure 7.5 shows that there are about 7 times more non-public schools than public Table 7.2 : National Exam Passing Rates, 2012-13 (In Percent) 2012-2013 Passing Rates Share of 6th Grade Students Sitting 6th 9th Rheto Philo For Exam* Ouest 77 77 30 37 70 Nord 82 79 22 35 58 Nord’Est 75 78 29 42 62 Nord’Ouest 78 71 21 28 61 Centre 70 82 32 46 56 Artibonite 79 71 26 35 60 Sud 77 65 23 33 52 Sud’Est 76 80 27 34 91 Grand’Anse 75 68 28 37 60 Nippes 72 70 28 47 58 National 77 76 29 38 64 *Note: Shares are calculated using 2010-11 School Census and 2010-11 exam results. Source: World Bank staff estimates using data from MENFP. 103 schools providing education at the primary been designated for primary education in recent level. A similar ratio also exists at the secondary years, and the major donors include the IADB, level. Non-public schools currently operate in CIDA, World Bank, WFP, UNICEF, UNESCO, a largely unregulated fashion, as the existing USAID, AECID, and AFD. Two of the largest licensing system has never been fully enforced. donor-financed programs are the TWP and the However, schools participating in certain public national school canteen program (Programme financing programs, such as the donor-financed National de Cantines Scolaires – PNCS). TWP Tuition Waiver Program (TWP), must comply finances a complete cycle of primary school with specific criteria for funding. for cohorts of children in disadvantaged communities by providing tuition waivers to 7.11. Today, Haiti has one of the world’s largest non-public schools, and PNCS funds school non-public sectors at the primary and meals and supplementation. These programs, secondary levels. According to the 2010-11 financed by multiple partners, supported over School Census (the most recent published), 94 one million students in the 2013-14 school year. percent of preschool, 78 percent of primary, and 73 percent of secondary students attend non- 7.13. Declining aid flows put into question, public schools.61 These shares are among the however, further progress and could even highest in the world. Slightly fewer than half threaten recent improvements. While primary of non-public primary schools are religiously enrollment rates have increased substantially in affiliated, with Protestant affiliated schools recent decades, enrollment is still not close to making up the majority of these. Another 35 universal. The most disadvantaged children – percent are privately owned schools, and the including the poorest, those living without their remaining 15 percent or so were founded and parents, and those with disabilities – remain out are run by communities. Among non-public of school. At the same time, declining donor secondary schools, 60 percent are non-religious financing and a recent decision by the Ministry and privately owned, and about 30 percent are of Education to stop funding tuition waivers religiously affiliated, with Protestant schools for new cohorts of first graders in non-public again making up the majority. While data from schools threaten the gains in access made in the 2012 ECVMAS show that poor children are recent years. more likely to attend public primary school, the majority of all children attend non-public C. … While Public Spending Has Been schools. Recently Rising … 7.12. International partners remain an important source of funding. As in many sectors in Haiti, 7.14. Public spending in education has been donor assistance for education has fallen since increasingly recently. Over the past decade, the large commitments made immediately after the budget of the Ministry of Education has the 2010 earthquake (Figure 7.6). However, at increased in line with the national budget and nearly 2 billion gourdes in 2013, it remains an accounted for about 14 percent of the total important source of funds for the education national budget (Figure 7.7). The budget of the sector. Between 80 and 90 percent of funds have Université d’Etat d’Haiti, a body independent of MENFP, has also increased in line with the national budget, accounting for 1-1.5 percent of 61 Preliminary results from the 2013-14 School Census the budget annually in 2012. Since the inception suggest that these shares have stayed fairly stable from of PSUGO (Programme de Scolarisation Gratuite 2010-11 to 2013-14. 104 BETTER SPENDING, BETTER SERVICES Box 7.1 : Education Provision in Haiti Formal Education in Haiti is structured in four levels (Figure below): • Preschool is meant to serve children from age 2 to 5, and is considered to have four levels based on these ages: poupons, petits, moyens, and grands. However, this structure is not formally mandated by public policy. • Fundamental education consists of three cycles. The first two cycles, consisting of 6 grades for children ages 6-11, are considered primary education. After these first 6 grades, children may enter into vocational programs, or continue to the third fundamental cycle (lower secondary) which consists of 3 grades for children ages 12-14. • Secondary: after lower secondary children may continue to secondary (upper secondary) education, which consists of 3-4 grades, depending on the model followed by the school, or enter vocational programs. • Superior (tertiary) education includes a range of university, technical, and vocational programs. Formal Education System in Haiti Preschool Fundamental Secondary Superior No formal structure Cycles 1 and 2: Cycle 3: Lower Upper Secondary Tertiary mandated by Primary secondary Goverrnment Ages 2-5 Ages 6-11 Ages 12-14 Ages 15-17 ou 18 Ages 18+ Grades 1-6 Grades 7-9 Grades 10-12 ou 13 Sources: MENFP (Ministère de l’Education et de la Formation Professionnelle) Figure 7.5 : Primary Schools, 1930-2011 Figure 7.6 : Donor Financing for Education Sector, (Thousands of Schools) 2010-13 (In Billions of Gourdes) 14 9 Non-public (R2011) 12 Public (R2001) 8 Non-public (R2003) 7 10 Public (R2003) 6 8 5 6 4 3 4 2 2 1 0 0 1930 1940 1950 1960 1970 1980 1990 2000 2010 2010 2011 2012 2013 Source : Sources : World Bank sta estimates using data from Sources : World Bank sta estimates using 2003 and 2011 Ministry of Finance (MEF ) School Censuses 105 Figure 7.7 : MENFP Annual Budget Funded by Domestic Resources, 2005-13 (Billions of Gourdes and Percentage of Total Budget (Rhs)) 8 16 7 14 6 12 5 10 Investments 4 8 Operations 3 6 Salaries 2 4 MENFP (% of total budget), rhs 1 2 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 Source : World Bank sta estimates using data from the Ministry of Economy and Finance (MEF) et Universelle) in 2011, public spending on 7.15. Overall, the addition of the tuition waiver education has, however, increased markedly. programs and school canteen program have In an effort to fulfill students’ entitlement to increased funding for primary education, and a basic education, the current administration likely improved the equity of spending. While introduced taxes on international phone calls the two tuition waiver programs, PSUGO and and money transfers, to be placed in a National TWP, as well as PNCS, are part of the investment Education Fund (Fonds National d’Education budget, they primarily consist of recurrent - FNE).62 Modeled after the donor-supported expenditures, and exclusively support primary Tuition Waiver Program (TWP), PSUGO schools. Each program is targeted in different provides tuition waivers for students in non- ways to more disadvantaged Haitians. Although public primary schools and fee waivers for 40 percent of students benefitting from PSUGO students in public primary schools through a are in the West, PSUGO and the Tuition Waiver per-student transfer made to schools. Other Program both provide waivers of USD90 per types of support are also provided to schools, but student, thereby self-targeting low-cost schools information on precisely what these are is not that serve poor areas (Figure 7.8).63 In addition, currently available. During the 2012-13 school schools were selected for participation in the year, over HTG 2 billion were spent on waivers, TWP to reach purposefully rural schools. The supporting approximately 1.4 million students PNCS has targeted schools located in areas of across over 10,700 schools (GoH 2014). nutritional vulnerability. 62 A US$0.05 fee is charged on every international phone call coming in and going out of Haiti, a US$1.50 fee is charged on every international money transfer. Although the FNE has not yet been approved by the 63 The waiver amount of $90 was set around 2007, based Senate, the funds have been channeled through the on earlier studies of the minimum costs of running a national budget (outside of the MENFP budget) to school in Haiti. New research is needed to understand finance PSUGO. how these costs have changed over time. 106 BETTER SPENDING, BETTER SERVICES 7.17. Even after the introduction of PSUGO, total Figure 7.8 : Student Bene ciaries by Department, public spending on education in Haiti is 2013 relatively low in a global context. Education (In Percent of Total) spending in Haiti is about 3.5 percent of GDP, below most LAC neighbors, as well as the global 100% Sud average for Low Income Countries (Figure 7.9). 90% 80% Sud Est However, this spending must be considered Artibonite relative to the share of the sector supported by 70% Centre 60% public finances. Of the roughly 3 million young Grand Anse 50% Nippes people in primary and secondary school in Haiti, 40% Nord about 50 percent are subsidized by PSUGO and/ 30% Nord Est or are attending public schools. While this marks 20% Nord Ouest 10% a substantial increase in the State’s role, a large Ouest 0% share of students remain without public support. Source : Gouvernment of Haiti 2014; World Bank sta estimates using data received from MEF. D. … Its Composition Is Not Conducive to Better Services … 7.18. MENFP’s operating budget is skewed towards non-“front line” staff. In 2013, for instance, 7.16. Implementation challenges have plagued almost 70 percent of the operating budget is PSUGO, however, and beneficiaries are likely allocated to salaries and personnel costs (Figure overestimated. In 2013, the government’s 7.10). This salary expenditure is in line with other anti-corruption unit (Unité de Lutte Contre la LAC countries, where 60-80 percent of the total Corruption -ULCC) published the results of education budget is absorbed by the wage bill an audit of about 30 percent of schools listed (World Bank, 2014). However, in these countries, as PSUGO participants. The report finds that the public sector pays the majority of all primary across all schools, 25 percent more students are and secondary teachers, while in Haiti the State reported to the program than actually exist, and only directly employs those in the small public in 11 percent of schools, there are major over- sector. Fewer than 400 inspectors are currently counts of at least 50 students (ULCC, 2013). The employed to serve the approximately 17,000 report also cites several cases of “ghost” schools: primary and secondary schools in the country, duplicated names and school information in and their salaries account for less than 1 percent the PSUGO records, schools that could not be of the wage bill. In addition to providing public physically located, and some cases of known fraud education, MENFP is meant to play a large role in (schools invented and registered in the program regulating and supporting schools, particularly for the private gain of specific individuals). In non-public schools. As a result, about 42 percent addition, the report raises several concerns of salary expenditures go to staff who are not about how the program is run, including dozens working as teachers, directors, or inspectors. of cases of over or under payment of transfers 7.19. Half of non-salary operating expenditures go to schools, lack of verification mechanisms, and towards textbooks, national exams, and the incomplete record keeping. national school canteen program. MENFP both provides free textbooks and subsidizes 107 Figure 7.9 : Public Expenditures on Education (In Percent of GDP) 8 7 6 5 4 3 2 1 0 Low Income Peru 2011 Guyana 2012 El Salvador 2011 Haiti 2012/13 Paraguay 2010 Chile 2011 Countries 2010 Ecuador 2012 Colombia 2011 Uruguay 2011 Mexico 2010 Argentina 2010 Brazil 2010 Aruba 2011 Jamaica 2010 Belize 2010 Bolivia 2011 Barbados 2010 Source : World Bank sta estimates using data from MEF; World Development Indicators students’ purchase of textbooks in public primary schools. This spending accounts for up Figure 7.10: Education Budget, 2014 to 24 percent of non-salary expenditures. A 2011 (domestic resources only, percentage of total) audit by the General Inspectorate of the Ministry of Finance (Inspection Générale des Finances – Services Consum. goods 13% 1% IGF) shows that approximately 66 percent of the Tangible textbook program budget went to subsidizing assets 1% books, while the remaining 34 percent went to Intangible assets providing free books (MEF, 2012). While over 2.5 0% Salaries million books were purchased, the audit found 69% Subsidies and quotas that only 19 percent of a sample of public schools 9% had actually received any books. IGF concludes Others 7% that the textbook program lacks a clear policy under which to operate (including on criteria Source : Ministry of Finance (MEF) for targeting beneficiaries, or requirements to return free books) and that management of the textbook program is insufficient, with little documentation available to show that books were ultimately given or sold at subsidized prices to students. The administration of national exams accounts for 14 percent of non-salary expenditures, and the operating expenses of PNCS account for another 9 percent.64 This 64 During the last years, there have been major investments on nutrition, health and school feeding 108 BETTER SPENDING, BETTER SERVICES leaves slightly more than 1 billion gourdes, or 14 percent of the operating budget, for all other Figure 7.11: Education Expenditures by Type, 2012 expenditures. (In Percent) E. … With Cost Remaining a Main 100% 90% Obstacle to Access… 80% 70% 60% 7.20. Total spending on education in Haiti is 50% relatively high. In total, approximately HTG 40% 30% 30-40 billion are spent on education in Haiti 20% annually, with the majority of funds coming from 10% 0% households. Estimates based on expenditure Poor Non Poor data from the 2012 household survey suggest Tuition Books Uniforms Transport Other that in total, households spend over HTG 20 Sources : World Bank sta estimates using ECVMAS 2012 billion per year on education at all levels in Haiti. In the past two years, the State has spent over HTG 10 billion per year, and international partners have disbursed HTG 2-3 billion. In total, this spending adds up to over 10 percent public schools, based on costs. Research using the of GDP. 2002-3 School Census finds that school fees are 7.21. Despite weak progression and learning, positively correlated with school infrastructure households spend a substantial amount (latrines, electricity), smaller class sizes, and on sending children to school. Among all more teaching materials (Demombynes, households with children age 6-14 in school, 93 Holland, and Leon 2010). Poor households in percent report positive education expenditures. particular spend relatively more on uniforms These expenditures are substantial: on average, and books, and less on transport, compared to these households reported spending 10 percent non-poor households (Figure 7.11). Against of total annual household consumption on this backdrop, costs remain a major obstacle education (including the education of all to education for poor households: when asked students in the household) in the 2011-12 school why school-aged children are not currently in year, and this share is equal for poor and non- school, 83 percent of households cite costs as the poor households. Because poor households primary reason (ECVMAS 2012) have more school-age children and lower total 7.22. For poor households, tuition spending over consumption, they spend less than half as much the past decade has increased less quickly than per child compared to non-poor households: for non-poor households, potentially due in HTG 3,600 compared to about HTG 11,400 per part to the growth in public financing. In 2001, 6-14 year old child per year. Households may the bottom 60 percent of households spent about therefore sort into schools, particularly non- HTG 450 per child for primary tuition; this had risen to about HTG 1,700 in 2012, for an programs funded by partners. The two largest school annual growth rate of 12.8 percent (IHSI 2003; feeding programs, funded and implemented by the ECVMAS 2012). In comparison, tuition for the World Food Program, the Canadian Government, and top 40 percent of households grew at nearly 18 the World Bank, covered nearly 600,000 children in 2014 at a cost of an estimated USD 30 million per year. percent per year, and overall consumer price 109 inflation averaged over 13 percent per year. This ratios, and are more likely to have a library and slightly slower growth in spending may be due electricity (Table 7.3). Across departments, in part to the growth of Government and donor student-teacher ratios are in between the financing in the sector. averages for LAC and Sub-Saharan Africa, while most departments have fewer electrified schools F. … And Disparities across Regions than the African average. While there appear to be a sufficient number of teachers, low levels and Poor Services of infrastructure across the country indicate that students are often lacking environments 7.23. Geographic disparities in service delivery conducive to learning. The physical proximity remain. The West department has more schools of schools also varies across the country. For per capita and generally better-resourced example, the number of schools per capita varies schools than other departments. While only substantially across communes, ranging from 1 14 percent of primary students in the West school (primary or secondary) up to 20 schools attend public schools, its large population and per 1,000 residents age 0-18 (Data sources: IHSI high enrollment rates mean that the West has 2012; 2010-11 School Census). Comparing the highest number of public primary students public to non-public primary schools, student- among all departments. In general, schools teacher ratios are substantially higher in public in the West, and in urban areas more broadly, schools, but measures of physical infrastructure appear to be better-resourced: primary schools and resources are similar on average. in the West have lower average student-teacher Table 7.3 : Characteristics of Primary Schools by Department, 2010-11 (In Percent) Percent of Schools with Share of Public Student-Teacher Canteen Library Electricity Enrollment Ratio (Units) Ouest 14 24.4 34 24 54 Nord 29 31.7 33 16 29 Nord’Est 31 36.8 46 14 16 Nord’Ouest 20 28.3 32 11 11 Centre 19 35.2 45 5 13 Artibonite 26 28.4 34 16 30 Sud 27 28.1 39 8 18 Sud’Est 31 29.1 36 14 16 Grand’Anse 31 32.3 26 8 8 Nippes 21 28.8 26 11 14 Sub-Saharan Africa 89 40.7 28 Latine America & 83 20.8 Caribbean Sources: UNESCO Institute for Statistics; World Bank EdStats, 2010-11 Haiti School Census. 110 BETTER SPENDING, BETTER SERVICES 7.24. The quality of service delivery in education several policy measures aimed at increasing is poor. Indicators of the quality of education, public accountability and improving quality in including teacher knowledge and learning the education sector. These measures include materials available at the school level, suggest mandatory registration of all teachers and that many children, particularly poor ones, schools with the MENFP to be instituted through are receiving a low-quality primary education. automatic, provisional granting of registration Instructional quality and the provision of cards followed by processes of checking learning materials are generally believed to be qualifications. In addition, new evaluations prior very limited (MENFP, 2013). For example, to the sixth grade will be instituted, and schools in French language and math assessments of performing particularly badly will be placed primary school teachers in the Central Plateau, under Ministry supervision. While these are where the questions were drawn from teacher welcome efforts, implementation and financing training institute exams, only 10 percent plans have not yet been articulated, and how (French) and 22 percent (math) of teachers these new plans fit with ongoing efforts is not were able to answer at least half of the questions clear. For example, how the new establishment correctly (MENFP and World Vision, 2013). identity cards for schools relate to the ongoing Low quality contributes to high repetition and effort to modernize the school licensing system, dropout rates, and ultimately to low educational which has been led by DAEPP (Direction d’Appui attainment, as children with weak basic skills are aux Ecoles Privées et Partenariat), has not yet unable to or gain little from completing primary been articulated. In addition, MENFP has been and continuing to secondary. planning to administer the Early Grade Reading and Math Assessments (EGRA and EGMA) 7.25. Efforts to increase accountability and improve to nationally representative samples of third quality, particularly in the large non-public grade students, and it is not clear how the newly sector, are growing, but much remains to be announced assessments will relate to these plans done. In August 2014, MENFP announced 111 112 BETTER SPENDING, BETTER SERVICES Chapter 8: Towards a more effective and better targeted social protection system 65 This chapter examines public expenditures on social protection, with a particular focus on social safety nets (SSNs). The review shows a positive increasing trend in SSNs spending, which corresponds to the government’s recent efforts to promote this area and implement new interventions under the umbrella of the Ede-Pèp Program. The overall SSNs system remains, however, fragmented, challenging the country’s weak administrative capacity. Going forward, the persistent low coverage and poor adequacy of interventions should be addressed, as well as institutional responsibilities clarified. Recent initiatives – such as the establishment of a national registry of beneficiaries and a national targeting tool – are key elements to improve the efficiency and equity of SSNs. A. Growing Investments in Social protection challenging (Grosh, 2004). These data challenges are exacerbated in a fragile and Safety Nets65 aid-dependent country like Haiti. Yet Haiti’s high 8.1. In many countries, attempting to estimate levels of poverty and vulnerability and a policy public expenditures in social protection is momentum for reform have prompted the need arduous. Lack of consensus on the scope of for such a review. Social protection programs and social protection, fragmentation of interventions policies, and especially social safety nets (SSN) across Ministries, and/or a weak or nonexistent - when correctly developed and appropriately sectoral strategy make a discussion of social funded - can directly address extreme poverty and help households both to manage shocks, 65 Prepared by Carine Clert, Frieda Vandeninden, and and invest in their children and their assets. In Victoria Strokova this context, a review of public spending in social 113 protection can help assess whether government’s strategic recommendations for social protection policy and programs are going in a direction that (gender equality, development of contributory contributes to objectives of poverty reduction and non-contributory mechanisms for workers and equity. and vulnerable groups). Similarly, some donor partners began to show interest in food and 8.2. The state has traditionally played a limited nutrition security, which led them to consider role in providing a safety net for the poor a potential safety net agenda that could address and vulnerable in Haiti. Observers (e.g both the short-term need to respond to constant Lamauthe-Brisson, 2013) have pointed out that shocks and the longer-term goal of reducing historically, Haiti never had a social movement chronic poverty. that challenged in any meaningful way the laissez-faire of the state on the specific issue of 8.4. An exploratory review of public spending on social protection. The Duvalier dictatorship safety nets during that period revealed that (1957-1986) left no room for such movements. despite progress, safety nets still remained Public social security organizations (insurance marginal. Existing programs at that time and social assistance) were created in the late were insufficient to meet the challenges of dire sixties but remained poorly developed. The poverty, constant exposure to weather shocks, post-dictatorship period saw the creation of new high food prices, and weak human development institutions such as the social investment fund indicators (Borgarello, 2009). Estimates of SSN FAES (Fonds d’Assistance Economique et Sociale) spending reached 0.7 percent of GDP (including in 1990. However, years of political instability (up donors’ contributions), which ranked Haiti in until the crisis of Aristide’s departure in 2004), the bottom quintile of low-income countries combined with the country’s constant exposure with low SSN spending. Lack of coordination to economic or natural disasters, contributed to of social protection policies and fragmentation a context of constant emergency and prevented among governmental entities also reflected the the development of long-term plans in which the absence of a national social protection strategy. provision of basic services and minimum social These conditions sparked partners’ interest in protection would have been essential elements. policy change. 8.3. The concept of social protection as a public 8.5. The January 2010 earthquake halted this policy that can contribute to poverty nascent interest in social protection and reduction and social risk management marked the beginning of three full years emerged progressively during the second that were understandably dominated by half of the 2000s. Marked by relative political humanitarian and short-term responses. and economic stability, this period saw the The earthquake affected only a portion of the emergence of some structural reforms. The country (mainly West and Southeast regions) 2007 National Growth and Poverty Reduction but caused over 200,000 deaths, considerable Paper (DSNCRP) covering the period of 2008- economic and infrastructure damage estimated 10 – the first poverty reduction strategy since at 120 percent of GDP. The government’s Action the democratic election of René Préval in 2004 Plan for National Recovery and Development – listed poverty reduction and the improvement of Haiti (PARDNH) included the establishment of living conditions as national objectives. This of a social protection system as a critical factor marked a milestone toward official recognition for recovery and growth in Haiti. This post- of the need to “protect” the poor. Emerging trade disaster plan focused both on short-term goals unions also approved a roadmap that included (for immediate reconstruction) and long- 114 BETTER SPENDING, BETTER SERVICES term objectives (until 2030). Since no safety PAARP also identifies a set of elements required net structure existed that could be rolled out to underpin the implementation of a social quickly, interventions had to be designed from protection system, such as a national targeting scratch. With the massive infusion of resources, system, a unique beneficiary registry, and an the earthquake both exposed and exacerbated integrated service delivery model at communal the deficiency in safety nets by multiplying the level through a network of multi-sectorial agents “project approach”, interventions, and actors. (along the Kore Fanmi approach),66 as well as local coordination on social protection (GOH 8.6. During the last three years (2012-14), there have 2014a and GoH 2014b). been encouraging signs of increased attention to the social sector agenda, and more recently 8.8. The national social assistance strategy “EDE to social protection as a framework to address PEP” represents the framework for several both short-term and long-term vulnerability flagship government programs. The goal of and poverty reduction needs. The institutional EDE PEP is to protect the vulnerable population apparatus was progressively rebuilt following living in extreme poverty throughout their life the earthquake, and efforts toward reform were cycle, in the short and medium term, to ensure resumed, including improvement of economic long-term investment in human capital and to governance and mid-term investment planning. provide opportunities to overcome the condition Education was made a top policy priority. In 2011, of extreme poverty. EDE PEP is based on four the Government launched a national program to complementary pillars: a) social inclusion, b) achieve universal access to basic education called development of human capital, c) economic PSUGO (Programme de Scolarisation Universelle inclusion, and d) development of a decent Gratuite et Obligatoire), which provides tuition environment. The key initiatives and programs waivers for students in public and non-public under EDE PEP are listed in Figure 8.1 below and schools (as discussed in Chapter 7). In the health detailed in Annex 4, which provides a glossary of sector, the Ministry launched a National Health programs. Strategy, an important attempt to improve 8.9. Most EDE PEP programs are financed, coordination of interventions, to strengthen the however, principally by concessional Ministry’s stewardship role, and to rally donors borrowing, which poses challenges in terms around key priorities. In 2012, the national of their sustainability. As illustrated Figure 8.1, strategy for fighting hunger and malnutrition, the source of funding varies according to each ‘Aba Grangou’, was also launched, with the aim program: PSUGO is financed by special taxes,67 of coordinating donor and government activities Kore Etidyan, Kore Ti Gran Moun, Restaurants in the nutrition sector. 8.7. A notable step toward the formation of a 66 Kore Fanmi (family coaching in Creole) is a social protection strategy was taken in 2014 Government initiative to link vulnerable families to with the launch of the three-year action plan available social services and provide them with some basic services and behavior change counseling. Both “Thinking and Fighting for a Haiti without the World Bank and UNICEF are supporting this Poverty: Action Plan for Accelerating the program. Reduction of Extreme Poverty” (PAARP). This 67 The PSUGO is financed through the National document provides concrete instruments for the Education Fund (Fonds National d’Education - FNE), implementation of the social protection system which is financed by taxes on international phone calls and defines a set of programs constituting a and international money transfers (collected mainly through two companies, CONATEL and BRH). See social assistance strategy named EDE PEP. The the chapter 6. 115 communautaires and PNCS by the national treasury, while the rest of EDE PEP programs are financed through concessional borrowing from Venezuela through the Petrocaribe agreement. As previously discussed, the recent decline in international oil prices is substantially reducing the availability of this source of financing , putting into question the sustainability of many of these social protection programs. Figure 8.1 : EDE-PEP SSN Financing, By Source, 8.10. Meanwhile, various donors are also 2013 showing signs of increasing interest in social (Percentage of Total) protection. The United Nations Development Public treasury Program (UNDP) has supported the human 34% Special capital approach of the newly established taxe conditional cash transfer Ti Manman Chéri. 45% The World Bank and United Nations Children’s Fund (UNICEF) are assisting the FAES with new models of support to and guidance of poor households in rural areas. The United States Agency for International Development (USAID) and the International Labor Organization Petrocaribe 21% (ILO) have been working with the Ministry of Social Affairs and Labor (MAST) and striving Sources : Sta estimates using FAES (2013), Primature (2014), to identify capacity-building needs. Efforts to Ministry of Education (MENFP), Ministry of Social A airs and Labor (MAST). coordinate with the government have also been Figure 8.2 : Programs Under EDE PEP EDE PÈP Social inclusion Human Capital Initiatives Environnement économiques Recurrent Periodic Health Education Kore Moun Panye Solidarité Family planing PSUGO Kore Peyizan Ranje Kay Katie Andikape Kantin Mobill Fight against School feeding Ti Credit HIMO Kore Ti Gran Moun Bon Solidarité cholera Literacy programs Community Community health Kore Etidyan Restaurants centers Ti Manman Cheri Carte Rose Notes: HIMO = labor intensive public works. PSUGO = Programme de Scolarisation Universelle Gratuite et Obligatoire. Source: GoH 2014a 116 BETTER SPENDING, BETTER SERVICES reinforced. For instance, the recent development 70 percent of rural households are considered of a national targeting system includes the chronically poor, compared with 20 percent in collaboration of different partners under the urban areas, highlighting the particular difficulty leadership of the government.68 of emerging from poverty in rural Haiti. Nationwide, only 14 percent of households are 8.11. The publication of poverty data, the first in transiently poor, and 12 percent of households a decade, marks an important milestone and reside above the poverty line but are deprived transformational opportunity for evidence- of access to basic services, suggesting that they based social protection.69 The EDE PEP lack the assets to improve substantially their framework, and in general any social protection well-being and remain out of poverty. Such a strategy in a given country, should ideally large share of chronic poverty, unusual in Latin respond to vulnerability and poverty needs. New America, highlights the structural challenges evidence is now available with the household of poverty reduction and limitations of the living conditions survey of 2012-13: this presents contributions of social protection interventions an historic opportunity to compare demand with without combined interventions around basic supply, and to assess whether the latter is up to services and income generation. the task. This is discussed in later sections of this chapter. 8.13. Haiti’s social protection system should also be able to include the transient poor. B. … Needed By A Vulnerable Haiti’s social protection system should also be capable of deploying short-term responses Population … given the constant exposure to shocks, which disproportionately affect the poor and the 8.12. The chronic poor should be an important extreme poor. With respect to the transient poor, focus of an effective social protection system one million people live slightly above the poverty in Haiti. Although poverty is widespread in line and could be pushed below by a shock. Haiti, as discussed in Chapter 1, rural areas and In terms of vulnerability to shocks, while a specific geographic departments require special typical Haitian household faces multiple shocks attention from the social protection system. annually, nearly 75 percent of households were Nearly half the households are considered chronically poor because they live below the moderate poverty line and lack at least three definitions for chronic and transient poor: Chronic poor are living below the poverty line and lack access of the seven basic dimensions defining non- to a basic set of services: education (the household monetary well-being (Figure 8.3).70 Almost head is literate; all school-age children are in school); health (the food security index); water (access to an 68 The committee in charge of the targeting tool, led by improved source of drinking water, including treated MAST, includes representatives from FAES, as well as water); sanitation (access to improved sanitation); principal donors such as USAID, UNICEF, UNDP, energy (access to a sustainable energy source); and WFP, the World Bank, and international NGOs such habitat (a home constructed of non-hazardous as CARE and Action Contre la Faim. The committee materials). Households deprived in at least three of aimed to develop a targeting tool that will respond to these dimensions are considered multi-dimensionally the specific needs of two major programs, Kore Fanmi poor or chronic poor (López-Calva, 2013). The and Kore Lavi (nutrition and food voucher program), transient poor lack monetary resources, but have as well as to serve the country more broadly by means access to basic services and are more likely to be able of a national tool. to move above the poverty line. In other words, they 69 World Bank (2015). are among the monetary poor but are not multi- 70 World Bank and ONPES (2014) uses the following dimensionally deprived. 117 Figure 8.3 : Chronic Poverty, Multidimensional Deprivation and Transient Poverty, 2012 (Percent of Total) 67 49 45 29 20 21 12 14 14 10 11 8 Chronic Deprived Transient Non-poor National Rural Urban Sources : ECVMAS 2012 and World Bank sta calculations. economically impacted by at least one shock in and other surveys heighten a potential role for 2012. This percentage rises to 95 percent for the social protection in bridging access gaps to extreme poor. Natural disasters, in particular, nutrition and social services that have a bearing have great disruptive potential partly because they on human capital. Financial constraints and have such a serious impact on agriculture, which other demand-side barriers to access and use of represents the main source of livelihood for most education and health services appear to be high of the poor, especially in rural areas. Indeed, the in Haiti, resulting in poor investment in human evidence shows that the most common covariate capital. As mentioned already in Chapter 1, the shocks are weather- or climate-related, while the main barriers to human capital accumulation most important idiosyncratic shocks are health- identified using the various data sources available related. This carries important implications are summarized below: for ensuring financial access to health services • The poor in Haiti suffer from malnutrition for the poorest as part of a broad-based social early in life and from subsequent protection strategy.71 food insecurity. According to the 2012 8.14. Poor households face tremendous constraints Demographic and Health Survey (DHS) one- to building, accumulating, and preserving fifth of children under five are chronically human capital although it should be malnourished. The national food insecurity potentially their biggest asset.72 Both ECVMAS rate was 38 percent in 2011 and 41 percent for rural areas (CNSA 2011). According 71 Covariate shocks affect large shares of the population/ to ECVMAS 2012, poor households were community (such as natural disasters or epidemics), much more likely to report repeated lack of while idiosyncratic shocks affect individuals (such as food or hunger when going to bed than non- sickness, death, or job loss). poor households. Importantly, households 72 Human capital is defined here broadly as a set of with young children (under the age of 5) are intangible assets, skills and knowledge that can create economic value and generate more valuable labor much more likely to experience repeated food outcomes. shortage. This is a particular cause of concern, 118 BETTER SPENDING, BETTER SERVICES as proper nutrition in early life is crucial where public interventions in social protection for brain development and subsequent life should go. The main findings are summarized outcomes (Alderman and King, 2006). below and key facts are detailed in Figure 8.4. The main policy implications – or demand in • The poor are much more likely to be illiterate social protection - ensuing from each key fact on (30 percent for the poor versus 15 percent for risks borne by the population throughout the life the non-poor) and to face lower levels of cycle, are summarized in Figure 8.5. education achievement (7 years of education for the non-poor versus 4.4 for the poor). • Poverty is highest among children and Poverty is also much higher for households relatively lower among adults. Almost 70 headed by persons with no education (77.6 percent of pre-school age children (below 6 percent of households headed by a person years old) live in poor households. While still with no education are poor). very high, poverty among adults is about 53.3 percent and almost half of all poor are adults • Poverty proves to be an important barrier to aged 18-64 years old. School-age children school enrollment. This is further supported represent about 30 percent of all poor, with the by the results derived from asking directly second highest poverty rate of 64.1 percent. why children are not in school: costs are Poverty rate among the elderly (above 65 cited as the primary reason by 83 percent of years old) is relatively lower, but higher than households (ECVMAS 2012). that for adults. Poor elderly represent the • The poor also face higher barriers to health smallest share of all poor (about 5 percent), in access. Financial barriers are the key obstacle line with their small share in total. The major to access to health care for the poorest. The risks are not attending school or dropping 2012 DHS pointed out that 76 percent of out of school for several reasons, particularly women age 15-49 declared that lack of money related to money or early pregnancy. deterred them from seeking health treatment. • A non-negligible number of children are • Finally, poor households are also more involved in child labor and many continue likely to resort to strategies that are to serve as “restavèks”, who work as domestic harmful to human capital accumulation. servants outside of their parents’ household. When there is a covariate economic shock Restavèks are difficult to identify in household in the community, a staggering 56 percent of survey data, yet some studies have shown the households in extreme poverty change their significance of this problem.73 nutritional behavior as opposed to 37 percent • Many adults (18-64 years old) in Haiti face of the non-poor households. The two events a risk of unemployment or not earning which are most likely to lead to the removal of sufficient income. This underscores the need a child from school are changes in household to reinforce the “promotive” features of a social composition (death or birth of a household protection system, as poor households need member) or decrease in monetary help from outside of the household. 8.15. The different risks supported by the 73 While ECVMAS 2012 includes only 91 observations population require different responses in terms with household members indicated as “Domestique of social protection policies. Looking at poverty = restavèk”, a 2009 study by the Pan American Development Foundation found, for instance, that headcounts by age group and risks across the life there may be as many as 225,000 restavèks living in cycle also offers important insights regarding Haiti (Pierre et al. 2009). 119 Figure 8.4 : Poverty Headcount by Age Group, 2012 (In Percent) 69.1 64.1 56.0 53.3 52.6 47.7 30.5 28.0 17.0 14.4 4.8 5.0 Child (<6 yo) School age (6-17 yo) Adult (18-64 yo) Elderly (65+ yo) Poverty headcount Share of poor Share of population Sources : World Bank Sta calculations based on ECVMAS 2012. to be connected to skill-enhancing and better main risk for the elderly is the lack of any income-earning opportunities. The principal pension (contributory or non-contributory risks facing adults are: unemployment, scheme) and/or access to healthcare, and the under-employment, low or variable income, dependency on family and charity for survival. informality, working but not earning enough Given demographic dynamics in Haiti, the income to cover basic needs (working poor), elderly tend to be somewhat neglected by unstable livelihoods and lack of access to antipoverty programs. As indicated above, physical and financial capital. the elderly represent less than 5 percent of the poor (Figure 8.4); however, poverty is still • Particular risks for women were highlighted. prevalent among this group as more than half The unemployment rate for women is twice as of those above 65 are poor. high as for men. The disparity is even higher in rural areas, where women are almost three times more likely to be unemployed than men. C. … But Too Fragmented … In rural areas, female-headed households have lower access to agricultural inputs such 8.16. The various programs building Haiti’s social as seeds, which can lead to lower productivity, protection system are excessively fragmented. thereby creating a gender gap. Within Government’s public spending on social protection, both the contributory (social • Youth (15-24 years old) were also found to insurance) and non-contributory pillars face additional challenges entering the labor (social safety nets) are spread across multiple market, with the lowest rates of participation institutions. While such dispersion may not and employment, and the highest rates of be unusual, it is problematic for a country unemployment and informal employment. with limited resources. Furthermore, this • Finally, the elderly (65 years and older) in fragmentation risks overwhelming the country’s Haiti are vulnerable to poverty and have limited fiscal and administrative capacity, and to rely on support from their families. The reducing the impact of public expenditures. 120 BETTER SPENDING, BETTER SERVICES Figure 8.5 : Key Risks, the Life Cycle, and Social Protection in Haiti Key facts on risks • Malnutrition: 22% su er from chronic malnutrition (DHS 2012) Young children (ages 0-5) • Mortality: Mortality rate is 92 deaths per 1000 live births, nearly 6 times higher than the regional average of 16 (WHO) Key implications • Structural programs aimed at malnutrition addressing supply and demand barriers • A more comprehensive early childhood policy Key facts on risks • Non-enrollment or school drop-out rates: approx. 200,000 children (ages 6-14) are estimated to School age children have dropped out of school (ECVMAS 2012) (ages 6-17) • Child labor: 20% of all children are involved in some form of work activity, 6% of children (ages 10-15) in the poorest quintile work and don’t attend school • Restavecs: up to 225,000 children are restavecs (PAHO 2009) Key implications • Structural social promotion programs to cover the direct and indirect costs (opportunity costs of child labor) of education for poor children Key facts on risks • Unemployment: 28.3% and up to 38% in urban areas and in Port-au-Prince Adults (ages 18-64) • Not earning adequate income: 80% of the rural poor live in households headed by a working person (61% in urban areas and 56% in Port-au-Prince) • Gender: Women are at a disadvantage in the labor market Key implications • Urban areas: temporary income generation programs that also incorporate improvement of training and skills-building • Rural areas: holistic programs for the extreme poor combining consumption smoothing and food security, productive projects and access to nancial capital • Address speci c constraints for women (care for children, the elderly or disabled) Key facts on risks • Lack of stable income: more than half of people over 65 are poor Elderly (ages 65+) • Lack of access to health care or other care Key implications • Targeted social pensions, solutions for access to health care or other care 121 Table 8.1 : Social Insurance: Institutions and Programs in Haiti, December 2014 Ministry and Institution Programs Target group MAST – ONA Private pension Travailleurs secteur privé MEF - DPC Public pension Civil servants MAST-OFATMA Public health insurance (since Oct. 2014/formerly Civil servants provided by Groupe Santé Plus) Disability insurance Private sector workers Work related insurance Private sector workers 8.17. Social insurance: Social insurance schemes than 20 programs, depending on the scope and are implemented by three different institutions, list of programs included. Given the objectives of depending on the program type (health and this paper, and as explained in the introduction, pensions) and work activity (public/private) this paper considers the following categories of (Table 8.1). Contributory schemes for private spending: (i) all social safety net programs under sector workers are managed under the general the EDE PEP flagship umbrella;74 and (ii) social authority of the Ministry of Social Affairs and subsidies for electricity and petrol. As illustrated Labor (MAST), but two separate institutions in Table 8.2 below, nine Ministries share oversight are responsible for the implementation of over different EDE PEP programs and 11 public private social insurance schemes, financed by agencies and/or Ministries are responsible for employers’ and employees’ contributions: (i) the their execution.75 In addition, the government Office of Work Accidents, Illness and Maternity provides universal subsidies for electricity and Insurance (OFATMA), which is responsible for petrol, which are managed by the Ministry of providing work injury and invalidity benefits Economics and Finance (MEF). Donor support to private workers; and (ii) the National Office is equally fragmented. for Old Age Insurance (ONA), which manages the private pension scheme. Regarding public D. … Too Small … sector workers, until recently health insurance and pension schemes were implemented by 8.19. Overall public expenditure levels for social two different institutions: (i) the Directorate protection in Haiti reached about 5 percent of for Civil Pensions (DPC) in the Ministry of GDP in 2013, when (i) contributions to social Finance (MEF), which manages pensions for insurance (1.2 percent of GDP); (ii) public civil servants and former military personnel spending on social safety nets (all EDE PEP or 1.4 and covered about 50,000 Haitian civil servants percent of GDP); and (iii) subsidies for electricity and their dependents in 2012 (300,000 people in total); and (ii) the private company for health 74 Definitions of EDE PEP programs are available in insurance of public sector workers, the Groupe Annex 8.1. Santé Plus. Since 2014, health insurance for both 75 Beyond EDE PEP, other small social safety net public and private workers has been consolidated interventions and ad-hoc initiatives have been under a single umbrella, OFATMA. identified, such as those implemented by the social assistance fund CAS (training centers) and the 8.18. Social Safety Nets (SSN): Government Institute for Social Welfare and Research (IBERS) spending on social safety nets is fragmented under the authority of the MAST. Other institutions across at least four Ministries and involves more are likely to finance SSN (e.g. BSEIPH and ONM) but data was not available at the time of this report. 122 BETTER SPENDING, BETTER SERVICES Table 8.2 : EDE PEP Fragmentation: Institutions and Programs by October 2014 Oversight Ministry Executing Ministry or EDE PEP Main target Groups Program (Ministère) de Tutelle) Public Agency Programs Type MAST Caisse d’Assistance Kore Ti Gran Moun Elderly Cash transfer Sociale (CAS) Kore Moun Andikape Elderly/Disabled Cash transfer FAES Kore Ti Gran Moun Elderly Cash transfer FAES / BSEIPH Kore Moun Andikape Disabled Cash transfer MAST Resto Communautaire n.a. Food FAES Resto Communautaire : n.a. Food Resto Pèp (Closed) FAES Panye Solidarite - - FAES Kantin Mobil n.a. Food OFATMA Carte Rose n.a. Access to Health Services FAES Kredi Ti Biznis (Ti Kredi) Women-headed hse Access to credit Ministry of Human Rights FAES Ti Manman Cheri Children Conditional Cash Transfer and Fight Against Extreme Poverty Ministry of Education MENFP School meals PNCS Students Food (MENFP) MENFP PSUGO (School Tuition Students Fee waivers Waiver Program) Secrétariat d’Etat à Literacy Programs Illiterate adults Training l’Alphabétisation FAES Kore Etidyan Students Cash transfer (scholarship) Ministry of Health MSPP Family Planning Women from 15 to 49 In kind (MSPP) years old Fight against Cholera n.a. In kind Commun. health centers n.a. In kind Ministry of the Promotion Kore Peyizan In kind of Peasantry (MARNDR) FAES • Kore plantè (seeds and Poor Farmers tools) FAES • Kore pechè (fishing Poor Fishermen equipment) MARNDR • Distribisyon betay Poor Farmers (distribution of livestock) Ministry of Women’s FAES Kredi Fanm Lakay (Ti Women in agriculture Credit Conditions (MCFDF) Kredi) Ministry of Interior FAES avec appui Direction Bon solidarité (bon dijans) Poor (emergency Cash transfer (MICT) Protection Civile situtation) Prime Minister’s Office. UCLBP (Unité de Ranje Kay Katie Poor Refurbishment of houses (Primature) Construction de Logement in poor neighborhoods et Bâtiments Publics) Ministry of Public Works SMCRS (Service HIMO (Labor intensive Unclear targeting. Self- Cash for work (MTPTC) Métropolitain de Collecte public works) targeting de Résidus Solides) Source: Staff based on Review of EDE PEP Official Documents (GoH 2014) and Key Informants’ Interviews 123 and petrol (2 percent of GDP) are included.76 However, SSN spending (Figure 8.6) is strikingly Figure 8.6 : Public Expenditure on Social Protection, 2013 low compared to that on universal subsidies that (Percentage of GDP) are regressive and fail to protect the poor and 3.0 vulnerable against shocks (Box 8.1). 2.5 8.20. Social Insurance. Social insurance contributions 2.0 (from employers and employees) represent 1.5 about 1.2 percent of GDP in 2013. Old age pensions for private sector workers is the largest 1.0 scheme, as contributions to pensions for public 0.5 sector workers, public health insurance and 0.0 disability benefits for private workers reach Social Insurance SSN Subsidies only 0.11 to 0.13 percent of GDP.77 Currently, Sources: World Bank sta estimates using data from Ministry about 400,000 workers are insured, yet only of Social A airs and Labor (MAST), Ministry of Finance (MEF), 3,000 retired receive benefits. Contributions to BOOST, FAES (2013) and Primature (2014). public pensions, including old age benefits for retired military, military staff and other civil servants have recently declined. The number of beneficiaries has remained stable since 2010, 8.21. Social Safety Nets. Government’s social safety with about 11,000 civil servants receiving old net spending has reached about 1.4 percent of age benefits. Contributions to health insurance GDP in 2013. Table 8.3 outlines spending data (public scheme) have also been steady over for EDE-PEP programs executed by MENFP, the last five years. Health insurance for private MAST and FAES for 2013.79 workers only provides work-related disability 8.22. This evolution represents an important benefits and maternity benefits but does not offer increase compared with past spending and health care insurance. Since 2012, OFATMA modifies the relative position of Haiti vis-à- has attempted to provide free health care to vis other low income countries in terms of specific target groups through the distribution SSN spending (Figure 8.7). Borgarello (2009) of insurance cards (‘Carte Rose’) on a pilot basis. estimated that SSN represented 0.7 percent However, no specific analysis of the experience is of GDP in 2008, including donor funding available to date.78 and in these estimates, government spending represented only 22 percent of total SSN. 76 Social insurance contributions do not represents the This recent increase in public SSN is mainly actual payments to insured beneficiaries (pensioners a consequence of the introduction in 2011 of and beneficiaries of health insurance). Actual the PSUGO fee waiver program and of other payments - for which data are not available - are EDE PEP programs starting 2013. In fact, financed out of contributions and therefore represent when analyzing the composition of EDE PEP, only a subset of overall contributions. 77 Contributions to public pension schemes correspond to government contributions only. additional data (including on the costs involved) is 78 The Carte Roz pilot program aims at evaluating the available. costs of the access to the expected health package. 79 According to the preliminary budget, other EDE-PEP In 2012-2013, about 400 health insurance cards programs (literacy, HIMO, cholera, community health were distributed and in fiscal year 2013-2014, the centers) were going to be implemented from 2013 or pilot extend to a group of university students but no 2014 on (GoH 2014a and Key informants’ interviews). 124 BETTER SPENDING, BETTER SERVICES Box 8.1 : Methodology Note on Social Protection Spending Given the multiplicity of actors, definitions of social protection, and lack of systematic data reporting, data collection for a public expenditure review on social protection is an arduous task. A comprehensive overview of spending on social protection requires disaggregated data spending (i) from the various social Ministries (MAST, MENFP, MSPP) and their different agencies; (ii) from the social insurance agencies within MAST and MEF (ONA, OFATMA, DPC and Groupe santé plus); (iii) from MAST agencies responsible for social safety nets (CAS and IBERS); and (iv) from the FAES (Fonds d’Assistance Economique et Sociale). As a result, the choice was made to focus primarily on 2013 and to present time-series data only when available (EDE-PEP programs) and feasible. Similarly, only major donors’ contributions to social safety nets (programs and technical assistance) are listed. 1. Social Insurance: (Contributions): • ONA and OFATMA (private schemes): data provided directly by the MAST • Public health insurance: data provided directly by MEF • Public pensions: BOOST database, World Bank 2014 2. Social Safety Nets: • EDE-PEP programs: FAES (2014), “EDE PEP, Programme nationale d’assistance Sociale: 1 an de mise en œuvre - 2 ans d’existence ”, Rapport 2013-2014 and FAES (2013), “EDE PEP, Programme nationale d’assistance sociale: Rapport 2012-2013, 1 an de mise en œuvre”, Rapport 2012-2013 • PNCS : data provided by MENFP • PSUGO : Primature de la République d’Haiti (2014), ‘Bulletin d’information Programme de Scolarisation Universelle Gratuite et Obligatoire’, Bulletin vérité, N°1, February 2014 • MAST Community restaurants: MAST (2013), ‘Rapport de suivi et d’évaluation du projet: programme de protection et d’assistance sociales sous-programme de restaurant communautaires’, Port-au-Prince, Novembre 2013 • SSN from CAS relating exclusively to programs mapped to EDE-PEP: data provided by MAST’s Directorate of Studies: (Kore Ti Gran Moun; Restaurant Communautaire). 3. Petrol and electricity subsidies: Data provided by MEF it is useful to distinguish between PSUGO as a context, spending on social safety nets has large tuition waiver program (representing 0.6 recently grown in comparison to that of other percent of GDP) and other EDE PEP programs low-income countries. (representing 0.8 percent of GDP).80 In this 8.23. In addition, donors’ interventions are widely fragmented, which compounds the difficulty 80 PSUGO is presented separately as its funding is of estimating exact contributions (Table 8.4). atypical. SSN executed by MAST include community In addition to the concessional borrowing that restaurants (restaurants communautaires) and an finances most EDE- PEP programs, other SSN aggregate figure for social protection for the elderly interventions are financed by various grants and vulnerable. The assumption was made that the latter refers to the social program for people with and implemented by external partners. Some disabilities under EDE-PEP, Kore Ti Gran Moun. examples include: the food voucher Kore Lavi 125 Table 8.3 : EDE PEP Executed Spending (Actual) in 2013 Program Title HTG In % In % of GoH Funding of GDP Expenditure Source PSUGO 2,229,097,353 0.61 2.94 Special tax Community Restaurants 550,000,000 0.15 0.73 Public treasury Panye Solidarité 490,673,688 0.13 0.65 Petrocaribe Kore Ti Gran Moun (MAST rogram for elderly 461,964,112 0.13 0.61 Public treasury and disabled) Kore Etidyan 443,322,200 0.12 0.58 Public treasury Ti Manman Cheri 265,507,530 0.07 0.35 Petrocaribe PNCS 235,000,000 0.06 0.31 Public treasury Bon Solidarité (bon dijans) 111,527,353 0.03 0.15 Petrocaribe Kore Peyisan 92,775,257 0.03 0.12 Petrocaribe Kantin Mobil 72,870,077 0.02 0.10 Petrocaribe Kore Moun Andikape 1,548,077 0.00 0.00 Petrocaribe Total 4,954,285,648 1.36 6.53 Source: FAES 2013, Primature 2014, MENFP, key informants’ interviews and WEO database. Figure 8.7 : Social Safety Net Spending – International Comparison, 2013 or latest (Percent of GDP) 4.0 3.5 3.0 2.5 2.0 1.5 0.61 1.0 0.5 0.75 0.0 KYRGYZ REP. AFGHANISTAN BENIN BANGLADESH TANZANIA NIGER MALI TOGO TAJIKISTAN CAMBODIA KENYA BURKINA FASO GAMBIA, THE MADAGASCAR RWANDA NEPAL HAITI LIBERIA MOZAMBIQUE ERITREA SIERRA LEONE SSN spending Spending on EDE PEP PSUGO Sources : World Bank sta estimates using dta from FAES (2013), Primature (2014), Ministry of Education (MENFP), Ministry of Social A airs and Labor (MAST) and IMF. 126 BETTER SPENDING, BETTER SERVICES Table 8.4 : Key Donors’ Contributions (Estimates) - Social Protection. Partners Budget in $US Period Budget per year ($US million) Multilateral USAID 80 millions 4 years (from 2013) 20,00 UNASUR 1 million 1 year (2013) 1,00 UNICEF 1.9 million 3 years (2012, 2012, 0,63 2014) World Bank 24.5 millions 2 years (2013-2014) 12,25 UNDP 1.7 million 3 years (2013-2015) 0,57 Caribbean Development NA NA Bk. Bilateral AFD 10 millions 2 years (2013 et 2014) 5,00 Canada NA Brazilian cooperation 3 millions 3 years (2011-2013) 1,00 Partial estimate of donors spending ($US million) 40,45 Partial estimate of donors spending (% of GDP) 0,46 Source: Key Informants program financed by USAID (budget of about Fanmi, implemented by FAES and financed by USD30 million, included in the social protection UNICEF (for $547,418 in 2012). budget of USD80 million over 3 years) and 8.24. Subsidies. The level of government spending on implemented by CARE; various school canteen subsidies has increased since 2007 and represents programs financed by multilateral organizations the largest share of the total GoH’s spending on (UNASUR, BDC) and by bilateral donors social protection. Government spending on (AFD and Canada) and operated by World subsidies represented almost 3 percent of GDP in Food Program (WFP), the World Bank and 2013. According to estimates from the Borgarello the donors’ Partnership Education for All, (2009), the surge in spending was initially driven which is coordinated by UNESCO; and other by the increase in electricity spending (from bilateral projects, such as milk distribution 0.84 percent in 2007 to 2.1 percent in 2012), financed by the Brazilian cooperation (budget as fuel subsidies actually decreased during the of USD3 millions) and implemented by WFP.81 period. A major concern, however, is the issue Partners also contribute to the financing of of regressivity of government subsidies. For some government programs with a technical instance, in the case of fuel subsidy: 95 percent assistance component: this is the case of the of the subsidy accrues to the richest quintile (see community agent (social worker) initiative Kore Chapter 9 for a more detailed discussion). E. … Not Well Targeted … 81 Donors and NGOs devote substantial resources to the school feeding programs. The World Bank and other 8.25. ECVMAS data points to the urgent need to donors are supporting the government in developing improve the targeting of social assistance a school feeding policy that would clarify roles, benefits, since as much as half of social responsibilities, accountability, objectives, financing, and would make the use of the resources more efficient (See Education Chapter). 127 assistance benefits accrues to the non-poor.82 The share of beneficiaries in the first two Figure 8.8 : Distribution of Social Protection Bene ts (Bene t Incidence), 2012 quintiles is less than half, but the size of transfers (Percentage of Total) tends to be bigger for higher quintiles, leading to 100 a much more regressive distribution of benefits 90 80 (Figure 8.8).83 This is especially the case for other 70 60 transfers,84 but also holds for food aid and, to 50 some extent, for scholarships. 40 30 8.26. A major finding of the most recent household 20 10 survey is that access to social security 0 All social Scholarship Other Food aid (contributory programs) is out of reach for assistance transfers most Haitians, especially the poor, leading Q1 Q2 Q3 Q4 Q5 to a lack of protection in old age or in case of Source : World Bank calculations using ADePT, based on sickness or disability. Only salaried employees ECVMAS 2012. working in the formal sector have access to the limited social insurance schemes existing in Haiti. Among the active population; i.e. those in the labor force, employees in wage employment majority do not (Figure 8.9). Access to social constitute only one-fifth, which corresponds to security is greatest among individuals in the less than 10 percent of the population. richest quintile. Two-thirds of employees with 8.27. Because of high levels of informality, only social security belong to the top quintile, while 11 percent of wage employees have access to only 5 percent are from the second poorest social security, primarily concentrated in the quintile. Given the prevalence of informality in upper quintiles of the population.85 Of wage rural areas, access is concentrated in the urban workers, only a small share (11 percent) have areas, particularly in Port-au-Prince. access to social security, while the overwhelming 8.28. Access to health insurance through employment in a firm registered in OFATMA 82 This share is reflective only of the programs captured is also very low. A very low percentage (less than in ECVMAS. It is currently not possible to estimate 4 percent) of the Haitian population has access what share of other assistance provided by the Government or NGOs goes to the poor. to health insurance administered by OFATMA. 83 Note that households are ranked on consumption net Most of the households with OFATMA belong to of social assistance transfers, so this is not simply as a the highest consumption quintile and live in the result of household moving up the quintiles due to the Metropolitan area. This is not surprising given transfers themselves. that OFATMA is only available to employees 84 The number of observations for other transfers is very of formal firms and their families, and receives small, so these estimates are less reliable. 85 For the purposes of this note, individuals employed social contributions from both employers and as wage earners who contribute to social security employees on a voluntarily basis (Cross et al.). or receive social security benefits, such as paid sick As discussed earlier, there could be potential leave or maternity/paternity leave, are considered to signs of future increases in the coverage of health have access to social security. The questions used for insurance, notably with the pilot introduction this analysis refer to individual employees. A set of questions referring to whether a firm is registered with of ‘Carte Rose’, a voluntary contributory health ONA or OFATMA provides similar levels of potential insurance which aims to extend coverage to the access to social security schemes. 128 BETTER SPENDING, BETTER SERVICES Figure 8.9 : Access to Social Security by Income Group, 2012 (Percentage of Total) Q1 Q2 25 0.3% 5.0% 20.9 Q3 7.8% 20 15 11.4 Q4 9.7 20.4% 10 4.6 4.8 5 Q5 0.4 66.5% 0 Q1 Q2 Q3 Q4 Q5 Total Source : World Bank calculations based on ECVMAS 2012 Source : World Bank calculations based on ECVMAS 2012 entire population, including informal workers. 2012 data, only about 8 percent of the Haitian However, the pilot seems still limited. population received non-contributory social assistance benefits such as scholarships, food 8.29. Because they lack access to contributory aid or other transfers in 2012 (Figure 8.10a).87 programs, poor Haitians have limited Overall, social coverage seems to be progressive: protection against poverty in old age or in about 12.5 percent of the population in the case of disability or sickness. Very few people poorest quintile receive some social assistance are eligible for contributory pensions when benefits compared with about 5 percent in the they retire, and those who are tend to be much two richest quintiles. A positive signal, given the better off. ECVMAS 2012 data shows that high prevalence of rural poverty, is that overall only 2.6 percent of the elderly (65 years and coverage is slightly higher in the rural areas, older) receive pensions (old age, disability) and primarily because of the larger share covered by the majority of them are non-poor. Pension beneficiaries overwhelmingly reside in urban areas (92 percent) and almost half (43.2 percent) extent of coverage of the EDE PEP programs, which live in the PaP metropolitan area. These results are not reflected in the ECVMAS. are consistent with the fact that access to social 87 Social protection programs are very fragmented security is very limited in rural areas. and often small in scale and coverage, so without surprise household survey data don’t capture many 8.30. Turning to SSN coverage, social assistance beneficiaries of such programs. This small number of coverage, is alarmingly low and well below observations should be kept in mind when interpreting the identified needs.86 According to ECVMAS the results. Preliminary findings from ECVMAS 2013 also confirm that overall coverage is low, on the order of 16 percent for social protection and about 13 86 This section uses two main methods to estimate how percent for social assistance, not including assistance much of the social safety nets actually go to the poor. from NGOs and religious organizations, coverage of First, it draws on the ECVMAS 2012. Second, it uses which is estimated at about 5.5 and 0.8 percent of the administrative data from FAES to shed light on the population, respectively. 129 Figure 8.10: Coverage of Social Assistance Programs and Distribution of Bene ciaries a. Population covered, 2012 b. Bene ciaries, urban vs rural, 2012 (Percent) (Percentage of total) 14 100% 12.5 11.7 12 11.3 80% 9.6 10 8.3 60% 8 7.1 6.8 6.0 5.4 40% 6 5.2 4.7 4.0 4 20% 1.7 2 0.9 0.9 0.5 0.9 0.3 0.3 0.5 0.4 0.4 0.3 0% 0.2 0 All SA Scholarship Food aid Other Poor Total Q1 Q2 Q3 Q4 Q5 transfers All social assistance Food aid Scholarship Other transfers Urban Rural Source : World Bank calculations based on ECVMAS 2012. Source : World Bank calculations based on ECVMAS 2012. food aid (8.8 percent compared to 5.3 percent in schools, such as free school meals (PNCS) or urban areas).88 More than 60 percent of food aid PSUGO, which are not captured in the survey. beneficiaries reside in rural areas (Figure 8.10b). EDE-PEP seems to lack sufficient focus on However, the beneficiaries of scholarships or early childhood. The plan for reduction of other transfers are slightly more likely to be in extreme poverty includes a few programs in the urban areas. early childhood window; however, most of the interventions (community restaurants, disaster 8.31. Social assistance coverage of various response programs, and health interventions) population groups is, nevertheless, not even: are insufficient and not tailored to the needs of young children are underrepresented among this age group. For example, health insurance is social assistance beneficiaries, which is a available only in urban areas and is contributory, concern given their vulnerability. Looking at so it is unlikely to reach the most vulnerable, the coverage by age group (Figure 8.11), children while community pharmacies do not focus on under 5 have the lowest coverage, with only preventative health and malnutrition, a critical 7.4 percent of all children under the age of 6 priority for young children. There is also little benefiting (indirectly) from social assistance available to prevent malnutrition or proactively benefits. This is of particular concern given that improve children’s development potential early this group suffers from the highest poverty rates, in life. Second, although child labor and restavèk as shown earlier. While coverage of school-age are a considerable phenomenon, this risk is not children is also quite low, they are much more addressed by the programs under the EDE PEP likely to benefit from programs targeted at strategy. Third, half of people above 65 are poor and coverage of contributive pensions is very 88 In contrast, more than half of the population benefits limited, yet the non-contributory cash transfer from remittances, which arguably play a role of an program only will be able to cover 30,000 people. informal safety net in Haiti. 130 BETTER SPENDING, BETTER SERVICES Figure 8.11: Coverage of Social Assistance F. … And of Limited Impact Programs by Age Group, 2012, 8.33. Unsurprisingly, ECVMAS indicates that (Percentage of Total Population) social protection programs have limited Elderly 8.5 9.7 impact on poverty and inequality because of 0.0 (65+ yo) 1.3 their low coverage and small value. Without 7.9 Adult (18-64 yo) 0.7 7.0 social protection transfers, including pensions, 0.1 8.0 poverty headcount would be less than one School age 6.8 (6-17 yo) 0.1 0.8 percentage point higher compared to the current Child (0-5 yo) 0.0 7.4 7.4 poverty rate. ECVMAS data also points out that 0.0 the value of most social assistance benefits (cash 0 2 4 6 8 10 or food) is very small and, hence, these benefits All SA Food aid Scholarship Other transfers contribute relatively little to the consumption Source : World Bank calculations using ADePT, of beneficiaries. With the exception of other Note : Direct and indirect bene ciaries transfers that are larger in absolute terms, other social assistance benefits are much less generous Figure 8.14, chart a). Scholarships, albeit small in absolute value, also contribute a large share 8.32. The geographic distribution of government (almost 37 percent) to the poorest quintile’s spending on its flagship, pro-poor safety net consumption (Figure 8.14 chart b). Overall, initiatives could be better aligned with the however, social assistance benefits contribute poverty incidence. Administrative data available only 10.6 percent to the poorest quintile’s from FAES on the extent of coverage for the last consumption. Contribution to consumption two years shows low coverage of some EDE PEP tends to decline as quintiles grow richer: for programs in regions with the highest poverty richer quintiles consumption is larger compared rates. The Centre region, with a poverty rate of to the value of benefits, so benefits are relatively 73 percent, has a similar coverage in terms of more important for the poorest. in-kind and in-cash transfer to the Ouest region 8.34. Design may also be revisited in some cases. (poverty rates of 37 percent). Similarly, Nord- While full analysis of the adequacy of all EDE PEP Ouest is the region with the highest prevalence of programs is not possible given data limitations, poverty (81 percent), but in-kind coverage is not there are concerns over adequacy of some higher than average. Coverage of cash transfers programs which deserve consideration. The CCT (Ti Manman Chéri, Kore Etidyan, Kore Moun Ti Manman Cheri brings into focus attention Andikape, Bon Solidarite, Bon Dijans) is lower to human capital by incentivizing mothers to than in-kind programs in every region (Figure keep their children in school. However, due to 8.12). Furthermore, meal distribution is not the budget restrictions and GoH decisions, the aligned with regional incidence of poverty, but is transfer is awarded to households only for a year. mainly concentrated in the capital. In 2012, the This puts into question its effect in the medium program of meal distribution Kantin Mobile was term, despite promising short-term effects that especially active in Port-au-Prince (Figure 8.13). are anecdotally reported at present. In this case, This is not surprising as this program focuses on the issue of adequacy is not just with respect to poor urban areas, to the detriment of rural areas the amount but the duration of the program. where poverty and food insecurity are relatively greater. 131 Figure 8.12: EDE PEP Coverage (FAES-Executed) by Region and Program Type, 2013 (Percentage of Total Population) 14 90 12 80 70 10 60 8.0 8 50 6 40 30 4 3.2 20 2 10 0 0 TOTAL OUEST ARTIBONITE SUD EST SUD NIPPES NORD CENTRE GRAND NORD EST NORD ANSE OUEST Cash In-kind Poor Source : data from FAES 2013 Figure 8.13: Numbers of Meals Distributed, Kantin Mobile, 2013 (Percentage of Total Population) 500 100 450 90 400 80 350 70 300 60 250 50 200 40 150 30 100 20 50 10 0 0 OUEST ARTIBONITE SUD EST SUD NIPPES NORD CENTRE GRAND NORD EST NORD ANSE OUEST Meals distributed (thousands) Poor (% of pop.), rhs Meals distributed (% of pop.), rhs Source : data from FAES 2013 Figure 8.14: Bene t Amounts and Contribution to Consumption of Bene ciaries a. Average annual per capita transfer by bene t type, 2012 b. Contribution to consumption of bene ciaries, 2012 (in Gourdes) (Percentage of total consumption per item) 35 8.3 6.2 Food aid 6.4 30 4.6 4.8 25 36.8 2.4 20 Scholarship 2.0 2.5 3.5 15 48.6 Other 63.8 10 14.6 transfers 5.8 5 49.0 10.6 0 All social 7.9 Scholarship Food aid. All SA Other Consumption assistance 6.6 transfers 4.4 6.4 Source : World Bank calculations using ADePT, based 0 10 20 30 40 50 60 70 on ECVMAS 2012. Q1 Q2 Q3 Q4 Q5 Source : World Bank calculations using ADePT, based on ECVMAS 2012. 132 BETTER SPENDING, BETTER SERVICES Table 8.5 : Alignment of EDE PEP Programs with Risks and Vulnerabilities Across the Life Cycle Life Cycle Risk Projects Under Planned Number of Stage ÉDE PÈP Beneficiaries (2016) 1. Early Childhood Malnutrition unaddressed n/a Mortality unaddressed n/a Poor child development unaddressed n/a 2. School age childhood Low school enrolment and drop-out PSUGO 1,500,000 School feeding 1,200,000 Ti Manman Chri 100,000 Child labor/Restavek unaddressed n/a 3. Youth Unemployment unaddressed n/a Poor educational outcomes Kore Etidyan 30,000 4. Adulthood Unemployment, lack of access to credit Ti Kredi 6,500 Female unemployment unaddressed n/a Low income, insecure livelihoods Kore Peyizan 100,000 Unemployment, low income HIMO (public works) n/a Poor living conditions, poor sanitation Ranje Kay Kartier/ Banm 25 districts Lumie- Banm Lavi Illiteracy Literacy 150,000 5. Old Age Low income Kore Ti Gran Moun 30,000 6. All Cycles Disability Kore Moun Andikape 30,000 Malnutrition and food insecurity Resto Communautaire 150,000 Natural disaster/emergency Panye Solidarité 600,000 Kantin Mobil 1,000,000 Bon Dijans n/a Disease/lack of access to health care Campaign for prevention n/a of cholera Community health n/a centers Carte Roz 2,500,000 Inadequate living conditions (lack of access Ranje Kay Kartier/ Banm n/a to sanitation, drinking water or waste Lumie- Banm Lavi but management) insufficient Violence unaddressed n/a Source: World Bank calculations based on data from FAES 133 134 BETTER SPENDING, BETTER SERVICES Chapter 9: Preventing a return of fuel subsidies89 This chapter discusses the distributional and price effects of adjusting fuel prices in Haiti. The rise in retail fuel prices in October 2014, combined with a decline in international oil prices has allowed Haiti to eliminate the fuel price subsidies that were burdening the budget. Such subsidies favor the rich, more so than subsidies on food, health or education. International oil prices are, however, volatile and without the introduction of an automatic price adjustment mechanism for petroleum products, fuel price subsidies could return to haunt the budget, but this time under tighter financing constraints. Fuel price subsidies are, however, only one dimension of broader development challenges: more generally, the overall policy framework for urban transport and renewable energies could be strengthened. A. Fuel Price Subsidies Implied a through adjustments in taxes. Taxes that should have been applied to petroleum products (e.g. Heavy Burden on the Budget…89 excises) were foregone and, when there were no 9.1. Like other net oil-importing developing more revenues to forgo, the government actively countries, Haiti has faced economic difficulties financed the difference in prices.90 Since 2010, the with high oil prices. Haiti has been for a long year of the devastating earthquake, until October time cautious passing through the increase in 2014, the retail prices for petroleum products international oil prices to consumers. In 2010, changed only once, in March 2011 (Figure 9.1).91 it abandoned a mechanism of automatic pass- through, and adopted a price control system 90 Appendix 5 presents the structure of oil products. 91 In October 2014, the government increased the 89 Prepared by Calvin Djiofack, Julie Saty Lohie and domestic price of oil for the first time since March Thiago Scot. 2011, as follows: about 7.5 percent for gasoline (to 135 Figure 9.1 : Evolution of Domestic Price and International Prices of Oil Products, 2004-14 (In Gourdes) 250 200 Gazoline 91 Gazoline 95 150 Diesel 100 Kerosene International Price 50 0 Aug-04 Jan-05 Juin-05 Nov-05 Avr-06 Sept-06 Fev-07 Juil-07 Dec-07 Mai-08 Oct-08 Mar-09 Aug-09 Jan-10 Juin-10 Nov-10 Avr-11 Sept-11 Fev-12 Juil-12 Dec-12 Mai-13 Oct-13 Mar-14 Oct-14 Sources : IMF and Government of Haiti 9.2. The size of oil subsidies in Haiti had increased have meant an effective price increase for fuels of dramatically over the last five years, imposing about 30 percent.93 a large burden on public finances. As the gap 9.3. Fuel price subsidies in Haiti were larger, relative between international and domestic prices to income, than most countries in the region increased, the cost of keeping retail petroleum and at a similar level of development (Figures prices fixed was placing an increasingly significant 9.3 and 9.4). While Haiti spent 1.5 percent of burden on public finances (nearly 2 percent of its GDP on oil subsides in 2011, the rest of GDP in 2013). Fuel subsidies in Haiti rose from Caribbean countries spent on average just 0.9 around 3.4 percent of total revenues in 2009 to percent of their GDP. The Dominican Republic, 11 percent in 2011, and to 15 percent in 2013 which shares the island with Haiti, spent less (Figure 9.2). In 2013, the Haitian government than 0.1 percent of GDP on fuel subsidies. The forewent revenue and spent 88.33 Gourdes, 30 amount of public resources allocated to fuel Gourdes, and 52 Gourdes per gallon to subsidize price subsidies in Haiti was also huge compared gasoline, diesel, and kerosene, respectively (Table to expenditures required to enhance growth 9.1).92 In early 2013, eliminating subsidies would and reduce poverty (Figure 9.5). Government 215 gourdes per gallon), 9 percent for diesel (to 177 price per gallon for gasoline, 16 percent for diesel and gourdes per gallon), and 9 percent for kerosene (to 171 23 percent for kerosene. gourdes per gallon). In February 2015 the increases 93 The simulations presented in this section were were reversed to reflect lower international oil prices, performed before the price changes of October 2014. with gasoline prices falling to 195 gourdes per gallon; All the baseline prices discussed in here, therefore, diesel to 157 gourdes per gallon; and kerosene to 156 refer to those fixed in the period between April 2011 gourdes per gallon. and September 2014. Average international oil prices 92 These subsidies represent about 31 percent of the total prevailing at the time were about USD 90/barrel. 136 BETTER SPENDING, BETTER SERVICES spending on fuel price subsidies had exceeded Figure 9.2 : Oil Subsidies Costs, 2010-13 public spending on health and education in (Billions of Gourdes and Percentage of GDP) recent years. 9.4. Furthermore, the price differential with the Dominican Republic may have encouraged 8 16 7 14 smuggling. The large difference in oil subsidies 6 12 between Haiti and the Dominican Republic had 5 10 created a strong incentive to smuggle petroleum 4 8 products to the higher-priced destination (the 3 6 Dominican Republic), increasing the budgetary 2 4 1 2 burden for Haiti. The price of gasoline in the 0 0 Dominican Republic was 30 percent higher, the 2009-2010 2010-2011 2011-2012 2012-2013 price of diesel 32 percent higher, and the price billions of gourdes % of revenues (rhs) % of GDP (rhs) of kerosene 24 percent higher, than in Haiti Source : Ministry of nance (MEF), 2014 Table 9.1 : Distribution of Oil Subsidies by Fuel Product, 2009-13 (Billion Gourdes) Fiscal Year Products 2010 2011 2012 2013 Gazoline 95 0.105 0.604 1.004 2.066 Gazoline 91 0.269 1.565 2.301 2.529 Gasoil 0.541 1.93 1.217 1.898 Kérosène 0.162 0.341 0.187 0.399 Total 1.078 4.44 4.709 6.892 Source. MEF (2014) Figure 9.3 : Total Post-Tax Costs of Petroleum Subsidies, Figure 9.4 : Total Post-Tax Costs of Petroleum Subsidies, 2011 2011 (Percentage of GDP) (In percent of General Revenues) 1.6 12 1.4 10 1.2 1.0 8 0.8 6 0.6 4 0.4 0.2 2 0.0 0 Dominican Low Income Caribbean World Haiti Dominican Low Income Caribbean World Haiti Republic Group Republic Group Sources : IMF (2014) and Government of Haiti Sources : IMF (2014) and Government of Haiti 137 Part of this difference is explained by take-up Figure 9.5 : Oil Subsidies and Public Spending rates: less than 3 percent of poor households Allocation, 2010-12 (Billions of Gourdes) report having spent any amount on fuel during the last 12 months, whereas almost 17 percent of 9 households in the richest decile did. 8 7 9.6. As a result, universal fuel subsidies are very 6 regressive. The poor receive only 1.6 percent of 5 total subsidies accruing directly to households 4 3 (Table 9.3 and Figure 9.7). Since direct usage of 2 petroleum products is so limited among the poor, 1 wealthier households benefit disproportionally 0 MENJS MSPP MAS MJSAC MCFDF Oil subsides from subsidies granted to these products. Over 2010 2011 2012 93 percent of fuel subsidies going directly to Source : Gouvernment of Haiti households benefit the richest 20 percent of the population. Poor households, on the other hand, receive only 1.6 percent of total subsidies, despite comprising 58 percent of the population. (IMF, 2014). Anecdotal evidence suggests that Extreme poor households benefit even less, informal trade was taking place. receiving a meager 0.3 percent of subsidies. 9.7. Fuel subsidies are particularly regressive when B. … Are Regressive… compared to subsidies on other categories of consumer expenditures, such as food, 9.5. Subsidized fuels are mainly consumed by the education and health. Expenditures on food wealthiest households. Poor households in Haiti represent more than 60 percent of the total consume very little fuel, both in absolute terms budget of households in the poorest decile, and as a share of their total budget. On average, but the share of food decreases steadily as poor households spend only 112 Gourdes a year expenditure level rises (Figure 9.8). Households on fuels, or 0.11 percent of their total annual in the richest decile spend only 40 percent of budget. 94 95 In comparison, the average annual their budget on food. The share of household expenditure on petroleum products in the expenditures on health is mostly stable across richest decile is 18,900 Gourdes, or 5.68 percent the consumption distribution, whereas the share of total consumption (Figure 9.6 and Table 9.2). of education increases slightly for the deciles 7 – 8 before declining for richer households. 94 All data discussed in this section, unless noted Thus the poor would receive between 30 percent otherwise, is from the 2012 “Enquête sur les Conditions and 40 percent of the benefits from a universal de Vie des Ménages après le Seisme” (ECVMAS 2012). reduction in education, health, and food prices The official poverty rates for Haiti are 58.5 percent (Figure 9.9). for moderate poverty and 24.4 percent for extreme poverty. 9.8. An increase in fuel prices could also have 95 ECVMAS 2012 does not have a specific question indirect effects on households. Directly, on expenditure on kerosene, which is widely used households are affected through their own in Haiti, especially by poorer and rural households. consumption of gasoline or kerosene. Indirectly, Despite not being able to quantify the expenditure on that product, we do provide a more qualitative they are also affected since petroleum products discussion on utilization later in this chapter. are used as intermediary products in many 138 BETTER SPENDING, BETTER SERVICES Figure 9.6 : Expenditure on Petroleum Products Per Decile, 2012 (Percentage of Budget) 6 5.68 5 4 3 2 1.46 1 0.46 0.41 0.08 0.03 0.07 0.09 0.18 0.14 0 Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10 Sources : ECVMAS (2012), World Bank and Ministry of Finance (MEF) calculations Table 9.2 : Average Yearly Expenditure on Fuels (HTG) Decile 1 40.05 Decile 2 22.29 Decile 3 61.70 Decile 4 95.46 Decile 5 221.88 Decile 6 197.19 Decile 7 753.99 Decile 8 746.97 Decile 9 3,097.85 Decile 10 18,887.00 Total 3,560.20 Source : ECVMAS (2012) and World Bank and MEF calculations. 139 Table 9.3 : Fuel Subsidies Received By Income Decile Figure 9.7 : Fuel Subsidies Received By Group, 2012 (Percentage of Total) (Percentage of Total) Poor Decile 1 0.10 2% Decile 2 0.00 Decile 3 0.10 Decile 4 0.20 Decile 5 0.60 Decile 6 0.50 Decile 7 2.10 Decile 8 2.30 Decile 9 10.80 Non-Poor Decile 10 83.20 98% Total 100.00 Sources : ECVMAS 2012, World Bank and Ministry of Finance (MEF) calculations Figure 9.8 : Composition of Expenditure by Figure 9.9 : Expenditures by Groups, 2012 Groups, by Decile, 2012 (Percentage of Total) (Percentage of Total Budget) 40.1 100 Decile 10 5.4 1.9 5.7 90 7.2 53.0 80 Decile 9 1.5 70 2.0 67,0 70,8 63,6 54.6 60 98,4 Decile 8 8.5 0.4 50 1.2 8.7 56.7 40 Decile 7 0.5 1.2 30 58.2 20 Decile 6 0.1 7.4 33,0 29,2 36,4 1.5 10 1,6 7.7 61.6 0 Decile 5 0.2 1.2 Fuels Education Health Food 7.6 61.9 Decile 4 0.1 Poor Non-Poor 1.9 7.5 63.1 Decile 3 0.1 Sources : ECVMAS (2012) and World Bank 1.9 7.7 64.6 Decile 2 0.0 1.6 7.0 64.9 Decile 1 0.1 1.6 0 10 20 30 40 50 60 70 Food Education Fuels Health Sources : ECVMAS (2012) and World Bank calculations 140 BETTER SPENDING, BETTER SERVICES Figure 9.10: Welfare Loss by Direct E ect (Percentage) 0.0 -0.2 -0.02 -0.01 -0.02 -0.02 -0.05 -0.04 -0.13 -0.12 -0.4 -0.6 -0.42 -0.8 -1.0 -1.2 -1.4 -1.6 -1.8 -1.65 Bottom Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Top Decile Decile Sources : ECVMAS (2012), World Bank and Ministry of Finance (MEF) calculations sectors and their higher price will feed into the richest two deciles.96 Since fuels are a very the price of the final good produced by these small part of poor households’ consumption sectors. To capture both these effects, the basket, the direct effect of a price increase is interdependence of sectors needs to be taken negligible (Figure 9.10). A 30 percent increase into account and a Social Accounting Matrix in fuel prices was assumed to assess the income (SAM) multiplier analysis has been used for distribution effects of this reform. Whereas this purpose. A SAM is an input-output table, the direct effect of eliminating subsidies would representing the transactions being carried out reduce the real consumption of the average poor between sectors for a given year. It describes all household by less than 0.05 percent, households the income received and expenditures made by in the richest decile would lose 1.65 percent. On households and various sectors, thus capturing average, a Haitian household can be expected to the linkages within the economy. Intuitively, our suffer a 0.62 percent drop in consumption due to model aims at estimating the loss of consumers’ the direct effect of higher fuel prices. purchasing power in real terms given a fixed budget and higher fuel prices. 96 The simulations presented in this section were 9.9. As discussed previously, the direct effect of a performed mostly using SUBSIM, The Subsidy rise in fuels prices affects almost exclusively Simulation stata package. For reference see Araar and Verme (2012). 141 Figure 9.11: Intensity In Oil Products By Sector Figure 9.12: E ect of Prices from the IO Table (Percentage of Total Oil Usage) (Percentage Increase) 40 35.0 33.5 14 13.3 35 30 12 11.3 25 20 10 15 8 10 5.0 4.0 5 2.1 1.9 1.7 1.1 0.9 0.8 6 0 4 2.6 Transports Electricity and utilities Chemical products Extractive Industries Metallurgical Other non-market services Electrical manufacturing Other market services Metal manufacturing Financial intermediation 1.5 2 1.1 0 Electricity and utilities Transportation Other market services Commerce Food Industry Sources : World Bank and Ministry of Finance (MEF) calculations Sources : World Bank and Ministry of Finance (MEF) calculations 9.10. Increases in fuel prices have large effects on electricity and transportation, however. The Figure 9.13: Contribution to In ation analysis of the indirect effects based the input- (Percentage Points) output matrix estimates the effect of increasing 0.7 fuel prices on each sector of the economy. An 0.6 0.5 increase in fuel prices would have the largest 0.4 direct impact on transportation (Box 9.1) and 0.3 electricity, in which more than 30 percent of total 0.2 input comes from oil products (Figures 9.11 and 0.1 9.12). Nonetheless, sectors that do not use fuel as 0.0 Transportation Long transportation Food Fuel Public services Personal care School Rent Cleaning Electricity a direct input can also be affected through input- output chains: even though the production of food is not intensive in fuel, its price increases more than one percent after a 30 percent rise in fuel prices because of the pass-through from Sources : World Bank and Ministry of Finance (MEF) agricultural and transportation prices. The overall calculations effect on the price level would nevertheless be contained: a 30 percent rise in fuel prices would lead to a one-time adjustment of about 3 percent in the overall price level (Figure 9.13). The main contributions to this increase would come from urban and interurban transportation (0.6 percentage points each), food (0.5 percentage points) and fuel (0.5 percentage points). 142 BETTER SPENDING, BETTER SERVICES Box 9.1 : The Transport Sector in Haiti Urban transportation in Haiti is dominated by private light vehicles and the “tap-taps”. As of 2005, the “tap-taps”, small buses or pickup trucks providing collective transportation, comprised almost 40 percent of the vehicle fleet in Haiti.97 Two-thirds of those used gasoline as their main fuel, but this number could be significantly lower nowadays, as differential taxation between gasoline and diesel has encouraged the use of the latter. Fuel costs are approximately half of total transportation costs, and changes are usually passed immediately to consumers, especially for price increases. Although transport prices are administered in Haiti, simulations using a Vehicle Operating Cost model (VOC) and assuming full pass-through suggest that a 10 percent increase in fuel prices would increase total transportation costs by 4 – 5 percent (4.7 percent for tap-taps).98The behavior of transporters in relation to prices seems, however, to be asymmetric: 61 percent of transporters surveyed in 2014 declared that they would raise tariffs if fuel prices increased, whereas only 43 percent said that they would pass price decreases to consumers (World Bank, 2014). Price increases of transportation affect mainly the richest households, but would perpetuate the lack of access of the poor to services, and may affect competitiveness. Expenditures on collective transportation are, on average, less than 2 percent of total budget for poor households, partly because many do not use the transport system. In that sense, increases in transport prices reinforce the exclusion of poorer households from transportation services and may hinder their access to economic opportunities. Tap-Tap Cost Structure Expenditure on transport, by quintile (percentage of total expenditure) Interest 6 26% 5 4 Carburant Depreciation 49% 3 0% 2 Crew 13% 1 Maintenance 0 Labor Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 2% Maintenance Lubricants Long transport Daily transport Total Tire Parts 1% 1% 8% Source : World Bank (2014) Source : Worl Bank sta calculations based on the 2012 97 Enquete Transport, IHSI (2005) 98 The Vehicle Operating Costs Model (VOC) is a stand alone program developed by the World Bank that estimates vehicle operating costs (See technical note on VOC Available at http://www.worldbank.org/transport/ publicat/wbtp-234.pdf). 143 There is considerable potential to improve energy efficiency and better manage fuel prices over the long term. Developing an efficient fuel stocking strategy is an important step towards reducing price volatility once subsidies are phased out. Guaranteeing lower prices also involves reducing the costs of internal fuel transportation. On the efficiency front, several measures could help reduce fuel waste: encouraging the renewal of the vehicle fleet, improving the conditions of roads and, in a more general sense, reducing transport needs by better urban planning. In the short term, targeted subsidies can be applied to compensate both transporters and consumers. International experience offers examples of targeted policies that reduce the costs of transportation for specific vulnerable groups: the “vale-transporte” for formal workers in Brazil; the free cable cars in Medellin and Rio de Janeiro; and other subsidies for students and the unemployed. Improving access to financing and reducing insurance costs could lower the costs faced by transporters and, therefore, final prices for consumers. 9.11. Poor households are especially vulnerable kerosene as the main source of lighting, since to increases in food prices. The modest electricity is available (with or without meters) (1.05 percent) rise in food prices nevertheless for only about 30 percent of households (Figures imposes a sizeable loss of welfare on the poorest 9.15 and 9.16). In rural areas, however, the use households, due to the large share of food in their of kerosene is much more prevalent. Among budgets (Figure 9.14). The lowest deciles are hit the rural poor, over 80 percent of households by a 0.7 percent decrease in real consumption use kerosene as main source of fuel. In urban due to food price increases alone, close to half of settings, reliance on kerosene is much lower, the total losses they suffer in the simulation. As with the exception of the extreme poor (bottom the level of income increases, higher food prices 25 percent). Thus, an increase in the price of loose in importance as a source of welfare loss, kerosene would be mostly felt by the most and increases in electricity and transportation vulnerable segments of the population. prices become more important,. 9.13. To sum up, the total effect of a fuel price 9.12. Furthermore, price increases of kerosene increase falls mainly on non-poor, urban would mostly affect poorer, rural households. households. Assuming a 30 percent in increase Even though there are no data on expenditures in prices of petroleum products, the consumption on kerosene, there is some indication that of non-poor urban households would decline poor households in rural areas would be the by more than 3 percent. Poor households face most affected by increases in the prices of this real losses between 1.5 – 2 percent, whereas product.99 Over 50 percent of households use households in the richest decile face decreases of over 4 percent in their real consumption 99 Even though ECVMAS 2012 provides information about expenditure on biomass (charcoal, wood…), ECVMAS 2012 does not provide details on different there is no survey question on kerosene expenditure. types of fuels consumed. Instead, the only question The survey does provide, however, information on the on fuel consumption refers to “Essence”, which can be main source of lighting for the household. Since the interpreted as any fuels used in vehicles. Therefore, the question is exclusively for the main source used, the scenario of a 30 percent increase in the fuel price refers data discussed here can be read as a lower bound for to a consumption-weighted average of increasing the number of households that consume kerosene. gasoline by 44 percent and diesel by 18 percent. 144 BETTER SPENDING, BETTER SERVICES Figure 9.14: Welfare Loss by Decile (Percent) Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7 Decile 8 Decile 9 Decile 10 0,0 -0,5 -1,0 -1,5 -2,0 -2,5 -3,0 -3,5 -4,0 -4,5 Fuel Food Transportation Long transportation Electricity Other indirect e ects Sources : ECVMAS (2012), World Bank and Ministry of Finance (MEF) calculations Figure 9.15: Main Source of Lighting, 2012 Figure 9.16: Households Using Kerosene As Main (percent Households) Source of Lighting, 2012 (Percent) Source % Rural Urban 1 Gas lamp (kerosene) 53.5 Electricity (no meter) 22.9 Utilize Kerosene (%) Candles 8.0 Electricity (individiual meter) 7.9 .5 Electricity (collectifve meter) 3.7 Other 1.9 Solar Panel 1.3 Generator 0.5 0 No lighting 0.3 0 20 40 60 80 100 0 20 40 60 80 100 Percentiles de depenses par tete Sources : ECVMAS (2012), World Bank and MEF calculations 145 Figure 9.17: Total Loss of Welfare by Decile Figure 9.18: Welfare Losses by Urban/Rural Areas (Percent) (In Percent) E et direct E et indirect Total 0.0 -0.5 Bottom Decile -0.02 -1.65 -1.67 -1.0 Decile 2 -0.01 -1.66 -1.67 -1.5 Decile 3 -0.02 -1.75 -1.77 -2.0 Decile 4 -0.02 -1.83 -1.85 -2.5 Decile 5 -0.05 -1.82 -1.87 -3.0 -3.5 Urban Rural Decile 6 -0.04 -1.92 -1.96 Decile 7 -0.13 -2.15 -2.28 Indirect e ects Direct e ects Decile 8 -0.12 -2.2 -2.32 Decile 9 -0.42 -2.33 -2.75 Top Decile -1.65 -2.74 -4.39 Sources : ECVMAS (2012), World Bank and Ministry of Finance (MEF) calculations Total -0.62 -2.27 -2.89 Sources : ECVMAS (2012), World Bank and MEF calculations (Figure 9.17). The average loss in consumption for oil subsidies. For a given price at the pump, is higher for urban households (more than 3 simple simulations illustrate, however, how percent) than for rural households (2.5 percent), changes in international prices could quickly because urban households spend more on fuel make fuel price subsidies return. The subsidies, and transportation than rural households do which amounted to around 2 percent of GDP (Figure 9.18). in 2013, would climb to over 3 percent of GDP with international prices at USD 100/barrel, but C. … And Could Return. drop to zero when prices hit USD 75/barrel. For oil prices lower than that threshold, current 9.14. The budget remains vulnerable to a rebound prices at the pump actually provides additional in international oil prices.100 In October 2014, revenues to the government. the authorities decided to increase petroleum 9.15. There is thus a need to return to an automatic prices, while international oil prices themselves pass-through. Lower international oil prices were starting to decline, eliminating the need have led people to the streets, demanding lower retail petroleum prices. The authorities in early 100 ECVMAS 2012 does not provide details on different 2015 have twice reduced retail prices. In the types of fuels consumed. Instead, the only question past, an automatic price adjustment mechanism on fuel consumption refers to “Essence”, which can be had been in place, but was abandoned in March interpreted as any fuels used in vehicles. Therefore, the 2010. Under this arrangement, any movement scenario of a 30 percent increase in the fuel price refers in international oil prices that would imply to a consumption-weighted average of increasing gasoline by 44 percent and diesel by 18 percent. a variation in retail fuel prices greater than 5 146 BETTER SPENDING, BETTER SERVICES Table 9.4 : Subsidies (Million Gourdes) Given Different Scenarios for Oil Prices Products / International Oil Gas Gasoil Kerosene Total Total (% GDT) prices (USD / barrel) 40 -4,898 -8,931 -1,848 -15,677 -4.3% 50 -3,000 -6,765 -1,426 -11,191 -3.1% 60 -1,101 -4,599 -1,005 -6,705 -1.8% 70 797 -2,433 -583 -2,219 -0.6% 80 2,696 -267 -161 2,268 0.6% 90 4,594 1,898 261 6,753 1.9% 100 6,493 4,064 682 11,239 3.1% Note : la simulation examine la relation entre les cours mondiaux moyens du pétrole et les prix à l’importation des produits pétroliers en 2012/2013, et procéder à une extrapolation linéaire dans les différents scénarios. percent was fully passed through. As a result, 9.17. More generally, the overall policy framework retail fuel prices had fluctuated more or less in for urban transport and renewable energies line with international oil prices. Consideration could be strengthened. Fuel subsidies are should be given to resume such a mechanism. only one dimension of broader development challenges. Thinking about a subsidy mechanism 9.16. With the uncertainty surrounding the future to contain transport price increases is an course of international oil prices, reducing or opportunity to discuss more broadly about the even eliminating the influence of politics in need to structure better the urban transport the setting of fuel prices would also be critical. sector with the aim of improving the security The implementation of the new automatic price of vehicles and the skills of drivers. Similarly, adjustment mechanism should be given to an discussions should be carried out on reducing independent agency in order to ensure, as much the reliance of Haiti’s economy on oil through as possible, that politics do not interfere with greater efficiency in the use of energy and the needed price adjustments. Such an agency could development of renewable sources of energy also be in charge of regulating the oil sector and (such as solar or wind). could include representatives from the ministries involved, importers, distributors, and possibly civil society. 147 148 BETTER SPENDING, BETTER SERVICES Concluding Observations 1. Human development indicators have greater financing, but the programs remain improved in Haiti over the past decade. Haiti fragmented and ill-targeted. In social protection, is the poorest country in the Latin America and the low coverage is staggering given the high Caribbean region, and among the poorest in the levels of chronic poverty, especially in rural world. The country has long been characterized areas. by its very low fiscal revenue mobilization, 3. Haiti needs to deal now with tighter budget seriously constraining its ability to carry out constraints. The decline in aid and in needed developmental spending (infrastructure, concessional financing is reducing the country’s health, education). Against this backdrop, greater fiscal space and putting the sustainability of its donor assistance and access to concessional higher level of public spending into question. financing has allowed Haiti to increase priority This is all the more worrisome as social sectors spending over the past decade. The fragmented are heavily depended on these resources for their fiscal data available would suggest that these financing. Within this context, the government additional resources have been channeled to faces an urgent need to use resources more the reconstruction and social sectors, consistent efficiently and equitably in order to have a with the decline in poverty and improvements greater development impact while keeping the in human development indicators (such as in fiscal accounts on a sustainable path. This study education) observed over the same period. aimed at contributing building an evidence base 2. Despite this progress, challenges remain. Haiti’s to inform decision making. economic system is not very inclusive. Finance 4. Against this background, some options going remains, one of the main obstacles for Haitians forward present themselves. While the report to access basic services in health or education. wanted to provide mainly a diagnostic and a When accessed, the service provided is often not better understanding of some critical sectors of the quality needed: problems of efficiencies such as health and education, some broad plague the sectors with health professionals options going forward can nevertheless be giving few visits and school teachers not using suggested as basis for discussion: their time efficiently. Social safety nets programs and education subsidies have benefited from 149 • Data: Overall tighter financing constraints and maintenance expenses (e.g., equipment, make it even more urgent to collect timely utilities), thereby limiting growth dividends data to track public spending appropriately of investment. and make it more efficient in different sectors. • Fiscal revenue: Balancing fiscal sustainability Differences in budget classifications and and development needs will remain a the absence of an effective Single Treasury daunting challenge for Haiti as long as the Account lead to a fragmentation of fiscal data country does not mobilize substantially more and prevents a comprehensive monitoring of own fiscal resources. In this regard, a review public spending from budget appropriation would be needed of existing exemption to payment. Fiscal reporting needs thus to be regimes for all tax instruments with priority strengthened to allow an appropriate tracking to customs duties and corporate income tax of public spending and set in place the where revenue losses are the greatest. An minimum conditions for transparency and important step would be the introduction of a accountability in the use of public resources. clear classification system in customs data to Completing the roll-out of the Single track the justification of exemptions. Where Treasury Account and unifying the budget tax exemptions or other investment incentives classifications among the various sources of are used, a clear framework needs to be put financing would be important steps in this in place to measure and track their costs. At direction. the same time, it is important to track and • Donors: With declining donor assistance, monitor the benefits these incentives were greater donor coordination is needed to make expected to deliver (such as employment the most of these resources. A master plan creation or technology transfer). for critical sectors translating the priorities of • Greater growth dividend: Haiti’s these sectors into areas of intervention could development needs are daunting and will federate the actions of the donor community, need to be addressed. In this regard, the allowing it to coordinate its technical and overall public investment management needs financial contributions, and avoid potential to be improved through a strengthening of duplication and inefficiencies. Such a project evaluation, selection, programming, plan would have indicators and deadlines, execution, and control. As a first step, the triennial operational plans, and an annual priority should be put on implementing action plans to structure the monitoring of its basic elements of formal project selection implementation. and appraisal, execution and controls such • Budget composition: The composition of as the need for investment projects to (i) public spending has room for improvement advance only if they are mature enough (i.e. to enable better service delivery. The growing developed in a concept or in a justification); wage bill, although in line with economies at (ii) be monitored (including physically); (iii) similar levels of development, has been the use resources adequately; (iv) are controlled driving force of total current spending, but ex-post as required. most of it is accounted for by staff not directly • Improve the performance of health facilities: involved in service delivery. Sharp increases For a long time the focus has been put on in capital outlays have not contributed to increasing the resources or the inputs. With the acceleration of growth, due in part to an shrinking budgets, the time may have come inadequate resource allocation for operating to shift the logic upside down and pay more 150 BETTER SPENDING, BETTER SERVICES attention to the results or the services actually assistance provide an opportunity to design a delivered by the health facilities through a less fragmented and better targeted system of more general use of results-based financing social safety nets. In the meantime, continuing (RBF). RBF could provide monetary and progress in the implementation of the social non-monetary incentives to health workers registry of potentially eligible beneficiaries is in order to increase their productivity, required. discouraging them from exercising another • Protecting the budget: With fiscal resources occupation. Health facilities may also have remaining limited, unbounded calls on the very low productivity because they have too budget should be contained. In this regard, few patients. To improve the use of their the re-emergence of fuel price subsidies resources, as well as health outcomes, such remain a risk. Higher international oil prices facilities may want to reinforce primary health could come back again to haunt the budget. level services and operate mobile clinics to There is thus a need to return to an automatic increase access to health services, particularly pass-through with this mechanism given to for the poor. an independent agency to reduce political • More accountable schools to increase influence in the setting of fuel prices. learning in the classroom: The majority of schools at all levels in Haiti are non-public, and operate with little oversight or accountability. With limited public resources, the sector will remain predominantly non-public for the foreseeable future and efforts should focus on making the schools more accountable for the quality of service they deliver. In this regard, efforts could focus on the comprehensive identification of schools or students benefiting from the different programs offered by the Ministry of Education, including the donor- financed Tuition Waiver Program (TWP) and PSUGO and linking this support with school achievements. Greater oversight could also be called for through a mandatory teaching license based on demonstrated competencies and a mandatory school identity card, leading to certification. In addition, targeted in- service training programs for teachers could be put in place to improve instruction in the classroom and learning. • Better targeting in social protection: Haiti’s social protection suffers from fragmentation of programs and interventions. The tighter resource constraints related to a reduced access to concessional borrowing and donor 151 152 BETTER SPENDING, BETTER SERVICES References Akazili J., M. Adjuik, S. Chatio, D. Kanyomse, A. Baez, J. and I. Santos (2007), “Children’s Vulnerability Hodgson, M. Aikins, and J. Gyapong (2008), to Weather Shocks: A Natural Disaster as a “What are the technical and allocative Natural Experiment”, Working Paper, The efficiencies of public health centers in World Bank, Washington D.C. Ghana?”, Ghana Medical Journal 42, no. 4, Barrientos, A. and D. Hulme (2008), “Social pp 149-155. Protection for the Poor and Poorest in Andrews, C., M. Das, J. Elder, M. Ovadiya and G. Developing Countries: Reflections on a Quiet Zampaglione (2012), “Social Protection Revolution”, Brooks World Poverty Institute in Low Income Countries and Fragile Working Paper, No. 30, Manchester, UK. Situations”, Social Protection and Labor Beeston, K. (2010), “Starting from Scratch: Building Discussion Paper No. 1209, The World Bank, a Brighter Future for Haiti’s Disabled Washington, D.C. Children”, June 14, 2010, The Guardian, Araar, A. and P. Verme (2012), “Reforming subsidies: www.theguardian.com. a tool-kit for policy simulations”, Guidebook Bellos, A., K. Mulholland, K. L. O’Brien, S. A. Qazi, M. and SUBSIM 1.0 manual, Policy Research Gayer, and F. Checchi (2010), “The burden of Working Papers No. 6148, The World Bank, acute respiratory infections in crisis-affected Washington D.C. populations: a systematic review.” Conflict Armington, P. (1969), “A Theory of Demand and Health 4, no. 3, doi:10.1186/1752-1505- for Products Distinguished by Place of 4-3. Production”, International Monetary Fund Bird, R., and P.P. Gendron (2011), The VAT in Staff Papers XVI, pp 159-78. Developing and Transitional Countries, vol.1, Backiny-Yetna, P., N. Garbiras and F. Marzo (2013), Cambridge University Press. “Poverty in Haiti: Methodological Note Borgarello, A. (2009), “Haiti Social Safety Net on the aggregate of consumption using Assessment”, World Bank Social Protection ECVMAS 2012”, mimeo. Non-lending technical Assistance Program, Mimeo. 153 Bourguignon, F. (2003), “The growth elasticity of Inequality Nexus - Fostering Shared poverty reduction: Explaining heterogeneity Prosperity”, Mimeo, The World Bank, across countries and time periods”, Inequality Washington, D.C https://editorialexpress. and Growth: Theory and Policy Implications, com/cgibin/conference/download.cgi?db_ edited by Eicher, T. S. and S. J. Turnovsky, pp name=CSAE2014&paper_id=381. 3-26, Cambridge, Massachusetts. Coa R., L.H. Ochoa and MEASURE DHS (2009) Brenton, P., C. Saborowski, C. Staritz and E. von Encuesta Nacional de Demografía y Salud, Uexkull (2011), “Assessing the Adjustment Bolivia ENDSA 2008, Macro International Implications of Trade Policy Changes Using Inc. Calverton, Maryland, United States. the Tariff Reform Impact Simulation Tool Coady, D., M. El-Said, R. Gillingham, K. Kpodar, (TRIST)”, World Trade Review 10, no. 02, pp P. Medas and D. Newhouse (2006), “The 249-276. Magnitude and Distribution of Fuel Bruhn, M. and J. Loeprick (2014), “Small Business Subsidies: Evidence from Bolivia, Ghana, Tax Policy, Informality, and Tax Evasion - Jordan, Mali, and Sri Lanka”, IMF Working Evidence from Georgia”, World Bank Policy Paper 06/247. Research Working Paper 7010, The World Coelli, T.J. (1996), A guide to DEAP: version 2.1: a Data Bank, Washington, D.C. Envelopment Analysis (Computer) Program. Buss, T. F (2013), “Foreign Aid and the Failure of Coordination Nationale De La Sécurité Alimentaire State Building in Haiti under the Duvaliers, (CNSA) (2011), Enquête Nationale de la Aristide, Préval, and Martelly”, WIDER Sécurité Alimentaire (ENSA). Working Paper No. 2013/104. Cros, M. and W. Zeng (2014), “Measuring Efficiency Carpenter, S., R. Slater and R. Mallett, Richard (2012), of Primary Health Care Facilities in Haiti “Social Protection and Basic Services in using Data Envelopment Analysis (DEA)”, Fragile and Conflict - Affected Situations” Accepted for a poster presentation at the Working Paper 8, Secure Livelihoods 3rd Global Symposium on Health Systems Research Consortium. Research in Cape Town, 30 Sept. - 3 Oct. Cayemittes, M., M. Fatuma Busangu, J. Bizimana, 2014. B. Barrère, B. Sévère, V Cayemittes and E. Demombynes, G., H. Peter and L. Gianmarco (2010), Charles (2013), Enquête Mortalité, Morbidité “Students and the Market for Schools in et Utilisation des Services, Haïti 2012, MSPP, Haiti.” Policy Research Working Paper 5503, IHE and ICF International, Calverton, The World Bank, Washington, D.C. Maryland, United States. Denizer, C., D. Kaufmann and A. Kraay (2011), Cellule de Planification et de Statistique (CPS/ “Good Countries or Good Projects? Macro SSDSPF), Institut National de la Statistique and Micro Correlates of World Bank Project (INSTAT/MPATP), INFO-STAT and ICF Performance”, Policy Research Working International (2014), Enquête Démographique Paper 5646, The World Bank, Washington, et de Santé au Mali 2012-2013, CPS, D.C. INSTAT, INFO-STAT and ICF International, Rockville, Maryland, United States. Devereux, S, (2000) Social Safety Nets for Poverty Alleviation in Southern Africa, A Research Christiaensen, L., C. P. Punam, and A. Sanoh report for DFID, ESCOR Report R7017. (2013), “Africa’s Growth, Poverty and 154 BETTER SPENDING, BETTER SERVICES Di John, J. (2010), “Development as State-making - FAES (2014), “EDE-PEP, Programme nationale Taxation, Resource mobilization and state d’assistance Sociale: 1 an de mise en œuvre – performance”, UKaid Crisis States Research 2 ans d’existence ”, Rapport 2013-2014. Centre Working Paper no. 84. FIAS (2010), “How to Reform Business Licenses”, Dollar, D., and A. Kraay (2002), “Growth is Good for World Bank Foreign Investment Advisory the Poor”, Journal of Economic Growth 7, no. Service, Washington D.C. 3, pp 195-225. Filmer, D. (2004), “If You Build It, Will They Come? Dollar, D., T. Kleineberg, and A. Kraay. (2013), School Availability and School Enrollment in “Growth Still Is Good for the Poor”, World 21 Poor Countries”, Policy Research Working Bank Policy Research Working Paper 6568. Paper 3340, The World Bank, Washington, D.C. Duryea, S., S. Galiani, H. Nopo, and C. Piras (2007), “The Educational Gender Gap in Latin Filmer, D. (2008), “Disability, Poverty, and Schooling America and the Caribbean”, Working Paper in Developing Countries: Results from No. 502, The Inter-American Development 14 Household surveys” The World Bank Bank, Washington, D.C. Economic Review 22, no. 1, pp 141-163. Easterly, W. and M. Woolcock (2006), “Social Ghana Statistical Service (GSS), Ghana Health Cohesion, Institutions, and Growth.” Service (GHS), and ICF Macro (2009), Working Papers 94, Center for Global Ghana Demographic and Health Survey 2008, Development, Washington, D.C. Accra, Ghana. Eaton, J. and S. Kortum (2001), “Trade in Capital Government of the Islamic Republic of Afghanistan, Goods”, European Economic Review 45, no. Ministry of Public Health (2014), Aid 7, pp 1195-1235. Coordination and SWAP Manual. Echevin, D. (2011), “Vulnerability and Livelihoods ______ (2010), Afghanistan National Health Accounts Before and After the Haiti Earthquake”, 2008-2009. Policy Research Working Papers, The World ______ (2003), A Basic Package of Health Services for Bank, Washington, D.C. Afghanistan, Transitional Government. Echevin, D. (2011), “Characterizing Poverty and Government of Cambodia, Ministry of Health (2009), Vulnerability in Rural Haiti: A Multilevel Cost and revenue comparisons for a sample Decomposition Approach”, MPRA Paper of contracted and non-contracted health 35659, http://mpra.ub.uni-muenchen. facilities, USAID, BASICS. de/35659/. Government of the Republic of Ghana (2011), “Joint Echevin, D. (2011), “Vulnerability to Asset Poverty Review of Public Expenditures and Financial in Sub-Saharan Africa”, MPRA No. 35660. Management”, Supported by the World Bank, http://mpra.ub.uni-muenchen.de/35660/. the Reconstruction Credit Institute (KfW) ECLAC (2013), Social Panorama of Latin America and DANIDA. 2012, United Nations: Santiago, Chile. Government of the Republic of Guyana, Ministry of FAES (2013), “EDE-PEP, Programme nationale Health (MOH), Bureau of Statistics (BOS), d’assistance Sociale: 1 an de mise en œuvre”, and ICF Macro (2010), Guyana Demographic Rapport 2012-2013. and Health Survey 2009, MOH, BOS, and ICF Macro Georgetown, Guyana. 155 Government of the Republic of Haiti (2014), “PSUGO Government of the Republic of Haiti, Ministry – Bulletin Verité No. 1”, Primature de la of Economy and Finance (2012), État République d’Haïti, Port au Prince, Haiti. d’exécution des dépenses budgétaires par institution et par secteur, exécutées à partir ______ (2014a), “Penser et Lutter vers une Haïti sans des fonds du Trésor Public d’octobre 2011 à pauvreté : plan d’action pour l’accélération septembre 2012. de la réduction de la pauvreté extrême”, http://www.mpce.gouv.ht/sites/default/files/ Government of the Republic of Haiti and Inter- parpsynthese.pdf. American Development Bank (2013), Rapport d’Evaluation du Système de Passation ______ (2014b), “Luttons pour une Haïti sans et Gestion des Marchés Publics d’Haïti. pauvreté: plan d’action pour la réduction de la pauvreté extrême”. Government of the Republic of Honduras, Secretaría de Salud, Instituto Nacional de Estadística ______ (2013), “EDE PEP: Stratégie Nationale (INE), ICF International (2013), Encuesta d’Assistance Sociale”, http://www.edepep. Nacional de Salud y Demografía 2011-2012, gouv.ht/4.03.2013.st rateg ieEDEPEP_ SS, INE and ICF International, Tegucigalpa, communication.pdf . Honduras. ______ (2001), Nomenclature des dépenses budgétaires Government of the Republic of the Ivory Coast, de l’État et guide d’utilisation. Ministère de la Santé et de l’Hygiène Government of the Republic of Haiti, Directorate- Publique. Comptes Nationaux de la Santé, General for Budget (2014), Budget de 2007-8, 2010. fonctionnement du Ministère de la Santé Government of the Republic of Kenya, Ministry of Publique et de la Population de 2003-2004 à Medical Services, Ministry of Public Health 2013-2014. and Sanitation (2010), Kenya National Health Government of the Republic of Haiti, Ministry of Accounts 2009-10. Planning and External Cooperation (MPCE) Government of the Republic of Rwanda, Ministry of (2012), Plan Stratégique de développement Health (2010), Rwanda Community-Based d’Haïti. Pays Émergent en 2030. Health Insurance Policy, Ministry of Health, ______ (2013), Grandes Réalisations, 2012-2013. Kigali . Government of the Republic of Haiti, Ministry Government of the Republic of Senegal, Agence of Public Health and Population (2014), Nationale de la Statistique et de la Rapport des Comptes Nationaux de Santé Démographie (ANSD), and ICF International 2011-12. (2012), Enquête Démographique et de Santé ______ (2013), Rapport des Comptes Nationaux de Continue (EDS-Continue 2012-2013), ANSD Santé 2010-11. and ICF International, Calverton, Maryland, United States. ______ (2013), Plan de Santé et de Développement d’Haïti 2012-2022. Government of the United Republic of Tanzania, Ministry of Health and Social Welfare (2012), ______ (2013), Plan Directeur de la Santé 2012-2022. National Health Accounts 2010. ______ (2012), Politique Nationale de la Santé. ______ (2012), Public Expenditure Review, Prepared ______ (2011), Rapport de la Carte Sanitaire du Pays, by Ministry of Finance and Development Programming and Evaluation Unit. Partners PER Macro Group. 156 BETTER SPENDING, BETTER SERVICES Grandcolas, C. (2005), “The Occasional Failure https://data.undp.org/dataset/Human- in VAT Implementation: Lessons for the Development-Index-HDI-value/8ruz-shxu. Pacific”, Asia-Pacific Tax Bulletin. Independent Evaluation Group (forthcoming), Grosh, M., C. del Ninno, E. Tesliuc and A. Ouerghi World Bank Group Assistance to Fragile and (2008), For Protection and Promotion: The Conflict-Affected States. An IEG Evaluation, Design and Implementation of Effective Safety The World Bank, Washington, D.C. Nets, The World Bank, Washington, D.C. Institut Haïtien de Statistique et d`Informatique Grosh, M. (2004), “Social Protection Treatment in (IHSI) (2012), Population Totale, Population Public Expenditure Reviews”, Guidance note, de 18 Ans et Plus Ménages et Densités Estimes The World Bank, Washington, D.C. en 2012, Ministry of Economy and Finance, Port au Prince, Haiti. Hanushek, E. and L. Woessmann (2009), “Do Better Schools Lead to More Growth? Cognitive ______ (2007), Projections de Population Totale, Skills, Economic Outcomes, and Causation”, Urbaine, Rurale et Economiquement Active, National Bureau of Economic Research Ministry of Economy and Finance, Port au (NBER), Working Paper No. 14633, NBER, Prince, Haiti. Cambridge, MA. ______ (2005), «Enquête sur le secteur transport». Harvey, P., R. Holmes, R. Slater and E. Martin (2007), ______ (2003), Enquête Sur les Conditions de Vie des “Social Protection in Fragile States”, Working Ménages (ECVH), Ministry of Economy and Paper, Overseas Development Institute, Finance, Port au Prince, Haiti. London, UK. Institut Haïtien de l’Enfance (IHE) and ICF Health Systems 20/20 Project. Democratic Republic of International (2014), Évaluation de the Congo (2011), National Health Accounts la Prestation des Services de Soins de 2008-2009, Health Systems 20/20 project, Santé (EPSSS) Haïti 2013, IHE and ICF Abt Associates Inc, Bethesda, MD. International, Rockville, Maryland, United Hernandez, A. R. and M. S. Sebastian (2014), States. “Assessing the technical efficiency of health Institut National de la Statistique et de l’Analyse posts in rural Guatemala: a data envelopment Économique (INSAE) and ICF International analysis”, Global Health Action, no. 7. (2013), Enquête Démographique et de Hidalgo, S. and M. Theodate (2011), “Inter-Agency Santé du Bénin 2011-2012, INSAE and ICF Real-Time Evaluation of the Humanitarian International, Calverton, Maryland, United Response to the Earthquake in Haiti”, Inter- States. Agency Standing Committee (IASC), Report. Institut National de la Statistique et de la Démographie http://www.alnap.org/pool/files/1395.pdf . (INSD) and ICF International (2012), Hollingsworth, B., P. J. Dawson, and N. Maniadakis Enquête Démographique et de Santé et à (1999), “Efficiency measurement of health Indicateurs Multiples du Burkina Faso, 2010, care: a review of nonparametric methods INSD and ICF International, Calverton, and applications.” Health Care Management Maryland, United States. Science, no. 2, pp 162-172. Instituto Nacional de Estadísticas y Censos (INEC) Human Development Index (HDI). “Value.” New York: and Ministerio de Salud (MINSA) (2002), United Nations Development Programme, Encuesta Nacional de Salud y Demografía 157 2001, Nicaragua and Macro International INTOSAI (2003), Assessment of Conditions Inc., Calverton, Maryland, United States. Encouraging Corruption, Vienna workshop, Bibliography. Institut de Statistiques et d’Études Économiques du Burundi (ISTEEBU), Ministère de la James, S. (2013), Tax and Non-Tax Incentives Santé Publique et de la Lutte contre le Sida and Investments: Evidence and Policy [Burundi] (MSPLS) and ICF International Implications, World Bank Investment (2012), Enquête Démographique et de Santé Climate Advisory Services. Burundi 2010, ISTEEBU, MSPLS, and ICF Josephson, E. and M. Bode (2013), Fiscal Space Analysis International, Bujumbura, Burundi. for Haiti, The World Bank, Washington D.C. International Finance Corporation (2011), “Economic Kauffmann, D. (2010), “Review of Cash-Transfer Integrated Zone, Market Analysis”, Coordination in Haiti Following the Washington, D.C. Earthquake of January 2010”, Cash Learning International Labor Organisation (2010), “La Partnership, http://www.cashlearning.org/ Promotion du Travail Décent dans la downloads/GroupURDpercent20-percen Reconstruction et le Développement d’Haïti t20Cashpercent20coordinationpercent20- après le Tremblement de Terre de 2010”, percent20Haiti.pdf. Geneva, Switzerland. Kauffmann, D. and O. Collins (2012), “Comparative International Monetary Fund (IMF) (2014), World Study of Emergency Cash Coordination Economic Outlook Database, April 2014. Mechanisms”, Cash Learning Partnership, http://reliefweb.int/sites/reliefweb.int/ ______ (2014), «Mécanisme de Tarification files/resources/120618percent20 Automatique des Produits Pétroliers”. Groupepercent20URD_Cashpercent20 ______ (2013), Optimizing Fiscal Policy for High c o ord i nat i on _ c omp ar at ive p e rc e nt 2 0 and Inclusive Growth in Haiti, IMF Country study_Final.pdf. Report 13/91. Kenya National Bureau of Statistics (KNBS) and ______ (2011), “Investing in Public Investment: An ICF Macro (2010), Kenya Demographic and Index of Public Investment Efficiency”, IMF Health Survey 2008-09, Calverton, Maryland. Working Paper IMF, Washington, D.C. Khwaja, M.S., R. Awasthi, and J. Loeprick (eds.) ______ (2008), “Haiti: Interim Poverty Reduction (2011), “Risk-Based Tax Audits - Approaches Strategy Paper”, IMF Country Report 08/115, and Country Experiences”, The World Bank, IMF, Washington, D.C. Washington D.C. ______ (2006), “Haiti: Interim Poverty Reduction Kingdom of Morocco (2014), “Revue des dépenses Strategy Paper”, IMF Country Report 06/411, publiques”, Report no. 86115-MA, The World IMF, Washington, D.C. Bank, Washington D.C. International Tax Dialogue (2005), “The Value Added Kirigia, J. M., L. G. Sambo, A. Renner, W. Alemu, Tax – Experience and Issues”, Background S. Seasa and Y. Bah (2011), “Technical Paper Prepared for the International Tax efficiency of primary health units in Kailahun Dialogue Conference on the VAT in Rome, and Kenema districts of Sierra Leone”, March 15–16 2005. International Archives of Medicine 4, no. 15. 158 BETTER SPENDING, BETTER SERVICES Kirigia, J. M, L. G. Sambo, and H. Scheel (2001), Situations” Working Paper 9, Secure “Technical efficiency of public clinics in Livelihoods Research Consortium. Kwazulu-Natal province of South Africa”, Marek, T., R. Eichler, and P. Schnabl (2006), “Allocation East African Medical Journal, no. 78, 1-3. de ressources et acquisition de services de santé Kleven, H.J. and M. Waseem (2013), “Using Notches en Afrique  : Qu’est-ce qui est efficace pour to Uncover Optimization Frictions and améliorer la santé des pauvres? ”, Série des Structural Elasticities: Theory and Evidence documents de travail sur le développement from Pakistan”, Quarterly Journal of humain dans la Région Afrique n°105. The Economics 128, no. 2, pp 669-723. World Bank, Washington, D.C. Kloeden, D. (2011), “Revenue Administration Marschall, P. and S. Flessa (2009), “Assessing the Reforms in Anglophone Africa Since the efficiency of rural health centers in Burkina Early 1990s”, IMF Working Paper 11/162. Faso: an application of Data Envelopment Analysis”, Journal of Public Health, no. 17, Kpodar, K. (2006), “The Distributional Effects of Oil pp 87-95. Price Changes on Household Expenditures: Evidence from Mali”, IMF Working Paper Ministry of Economy and Finance (MEF) (2014), 06/91. “Evaluation des effets de la réforme des subventions aux carburants et ses Lagomarsino, G., A. Garabrant, A. Adyas, R. Muga, implications”, Presentation at the Second and N. Otoo (2012), “Moving towards Workshop on Options for Fuel Price Reform, universal health coverage: health insurance May 2014, Port-au-Prince, Haiti. reforms in nine developing countries in Africa and Asia.” Lancet, no. 380, pp 933–43. Ministry of Economy and Finance (MEF), Inspectorate General of Finances (2012), Lamauthe-Brisson, N. (2013), “Social protection Audit du Programme de Subvention et de systems in Latin America and the Caribbean: Dotation de Manuels Scolaires 2009-2010, Haiti.” Project Documents, Economic Rapport Final, MEF, Port au Prince, Haiti. Commission for Latin America and the Caribbean and United Nations. Ministry of National Education and Vocational Training (MENFP) (2013), Programme Lombardo, A. (2012), “Mapping of Social Protection D’Interventions Prioritaires en Education Programmes in Haiti”, Oxfam, UK and (PIPE) : 2013-2016, MENFP, Port au Prince, UNICEF. Haiti. López-Calva, L.F., D. Battiston, G. Cruces, A.M. Ministry of Planning and External Cooperation Lugo, M. Santos (2013), “Income and (MPCE), Direction de l’Investissement Beyond: Multidimensional Poverty in Six Public (2013), “Analyse de l’Exécution Latin American Countries”, Social Indicators du Programme d’Investissement Public Research, Springer, vol. 112, no. 2, pp 291- 2012/2013”. 314. Ministry of Public Health and Population (MSPP) Lunde, H. (2008), Youth and Education in Haiti: (2013), Rapport des Comptes Nationaux de Disincentives, Vulnerabilities, and Constraints, Santé 2010-11. FAFO, Oslo, Norway. Ministry of Public Health and Population (MSPP), Mallet, R. and R. Slater (2012), “Growth and Haitian Child Institute, and ICF International Livelihoods in Fragile and Conflict-Affected (2013), “2012 Haiti: Mortality, Morbidity, 159 and Service utilization Survey: Key Findings”, Child Trafficking, Restaveks, and Victims Final Report, MSPP, IHE, and ICF, Maryland, of Violence, Pan American Development United States. Foundation, Port-au-Prince. Mohapatra, S., B. Moreno-Dodson and D. Ratha Primature de la République d’Haïti (2014), “Bulletin (2012), “Migration, Taxation, and Inequality”, d’information Programme de Scolarisation World Bank Economic Premise Note 80, The Universelle Gratuite et Obligatoire”, Bulletin World Bank, Washington, D.C. vérité no. 1. Moreno-Serra, R., S. Thomson, and K. Xu (2013), Rajaram, A., T. Minh Le, N. Biletska, and J. Brumby “Measuring and comparing financial (2010), “A Diagnostic Framework for protection”, Health system performance Assessing Public Investment Management”, comparison, an agenda for policy, information WPS5397, The World Bank, Washington, and research, edited by Irene P. and P. C. D.C. Smith, pp 223-255, Open University Press. Ravallion, M., and S. Chen (2007), “China’s (Uneven) Morgan, L. (2012), “Results-Based Financing for Progress against Poverty”, Journal of Health: Performance Incentives in Global Development Economics, 82, no. 1, pp 1-42. Health — Potential and Pitfalls”, The World Rawson, I. and G. Berggren (1973), “Family Structure, Bank, Washington D.C. Child Location and Nutritional Diseases in Moser, C. (2006), “Asset-Based Approaches to Poverty Rural Haiti.” Journal of Tropical Pediatrics Reduction in a Globalized Context”, Global and Environmental Child Health, no. 19, pp Economy and Development Working Papers, 288-298. The Brookings Institution, Washington, D.C. Research Triangle Institute (2010), Haiti  : Early Moser, C. (1998), “The Asset Vulnerability Grade Reading Assessment - Rapport pour Framework: Reassessing Urban Poverty le Ministère de l’Education et la Banque Reduction Strategies”, World Development Mondiale, RTI : Research Triangle Park, NC. 26, no. 1, pp 1-19. Sachs, J. (2001), “Investing in Health for Economic National Bureau of Statistics (NBS) [of Tanzania] and Development: Report of the Commission on ICF Macro (2011), Tanzania Demographic Macroeconomics and Health”, Commission and Health Survey 2010, NBS and ICF Macro, on Macroeconomics and Health, Geneva. Dar es Salaam, Tanzania. Saksena, P., K. Xu, and V. Durairaj (2010), “The drivers Niyuhire, F (2010), “More Health for Every Dollar: of catastrophic expenditures: outpatient Results-Based Financing”, The World Bank, services, hospitalization or medicines?” Washington D.C. World Health Report, Background Paper no. 21. OECD (2009), “Taxation, State Building and aid”, OECD Factsheet-Update. San S., M., and H. Lemma (2010), “Efficiency of the health extension programme in Tigray, Pan American Health Organization (PAHO) (2010), Ethiopia.” BMC International Health and Haiti Health Facilities: Monitoring Health Human Rights 10, no. 16. Service Response Capacity, http://ais.paho. org/phip/viz/haiti_healthfacilities_v2.asp . Schieber, G., C. Cashin, K. Saleh, and R. Lavado. (2012), “Health Financing in Ghana”, Pierre, Y. F., G. R. Tucker, and J-F.Tardieu (2009), “Lost Childhoods in Haiti: Quantifying 160 BETTER SPENDING, BETTER SERVICES The World Bank, Washington, D.C., for-innovative-international-financing-of- doi:10.1596/978-0-8213-9566-0. health-systems/. Scheiman, I., J. Langenbrunner, J. Kehler, C. Cashin, UN Special Envoy for Global Education (2013), and J. Kutzin (2010), “Sources of funds and Accelerating Progress to 2015 Haiti. revenue collection: reforms and challenges”, UNESCO, UNESCO Institute for Statistics Implementing health financing reform: lessons database, UNESCO, Paris, France, from countries in transition, edited by J. http://data.uis.unes co.org/Index. Kutzin, C. Cashin, and M. Jakab, p. 87-118, aspx?DataSetCode=EDULIT_ The World Health Organization, Geneva. DS&popupcustomise=true&lang=en Unité Secure Livelihoods Research Consortium (2012), de Lutte Contre la Corruption (ULCC). “Social Protection and Basic Services in 2013. Rapport de Vérification du Programme Conflict-Affected Situations: What Do We de Scolarisation Universelle Gratuite et Know?”, Briefing Paper 2, ODI, London. Obligatoire (PSUGO). ULCC: Port au Prince, Haiti. SEDLAC (2014), Socio-Economic Database for Latin America and the Caribbean – CEDLAS ______ (2014), Education For All Global Monitoring and the World Bank, Enrollment in LAC Report 2013/14, UNESCO, Paris, France. Database. UNICEF (2011), Children in Informal Alternative Shepard, D., D. Hodgkin, and Y. E. Antony (2000), Care, Discussion Paper, UNICEF, New York. Analysis of hospital costs: a manual for ______ (1999), Child Domestic Work, Innocenti managers, The World Health Organization, Digest, International Child Development Geneva. Center, Florence. Street, A., J. O’Reilly, P. Ward, and A. Mason USAID (2012), The Early Grade Reading Assessment (2011), “DRG-based hospital payment and in Haiti, USAID: Washington, D.C. efficiency: Theory, evidence, and challenges”, Diagnosis Related Groups in Europe. Moving ______ (2011), Health service delivery costing. Health towards transparency, efficiency and quality in center analysis, Management Sciences for hospitals, edited by R. Busse, A. Geissler, W. Health (MSH). Quentin, and M. Wiley, pp 93-114, European ______ (2009), Costing the basic package of health Observatory on Health Systems and Policies services in Liberia at clinics and health centres Series, WHO Regional Office for Europe, in Liberia. Rebuilding Basic Health Services Copenhagen. Project (RBHS), JSI Research and Training Suzuta, E. (2011), “Education in Haiti: An overview Institute, Inc., the Johns Hopkins Center of Trends, Issues and Plans”, WISE Qatar for Communications Programs (JHUCCP), Foundation, http://www.wiseqatar.org/sites/ JHPIEGO, and Management Sciences for default/files/files/4851/eduoverview0921_ Health (MSH. draft_with_cover.pdf Verner, D. (2008), “Labor Markets in Rural and Urban Taskforce on innovative funding international Haiti: Based on the First Household Survey financing for health systems, for Haiti”, Policy Research Working Paper 4574, The World Bank, Washington, D.C. http://www.internationalhealthpartnership.net/en/ about-ihp/past-events/high-level-taskforce- Wagenaar, A. C., M. J. Salois, and K. A. Komro (2009), “Effects of beverage alcohol price and tax 161 levels on drinking: a meta-analysis of 1003 Public Expenditure Review, Africa Region, estimates from 112 studies” Addiction, no. The World Bank, Washington D.C. 104, pp 179–190. ______ (2008), Haiti: Public Expenditure Management World Bank (2014), Haiti Poverty Assessment: and Financial Accountability Review, The Poverty Profile and Trends, The World Bank, World Bank, Washington, D.C. Washington D.C. ______ (2006), Haiti: Social Resilience and State ______ (2014), Résultats de l’évaluation douanière Fragility in Haiti. Report No. 36069, The (CAT‐R) préparés pour le Service des douanes World Bank, Washington, D.C. de Haiti, The World Bank, Washington, D.C. ______ (2004), Benin Enhancing the Effectiveness of ______ (2013), “Strengthening Competitiveness Public Spending. A Review of Three Sectors, and Promoting Diversification in Haiti - The World Bank, Washington D.C. Diagnostic Trade Integration Study”, The ______ (1998), Haiti The Challenges of Poverty World Bank, Washington, D.C. Reduction, Report No. 17242-HA, Volume I, ______ (2013), Stop Conflict, Reduce Fragility and End The World Bank, Washington, D.C. Poverty: Doing Things Differently in Fragile ______ , World Bank EdStats: Education Statistics and Conflict-affected Situations, World Bank (database), The World Bank: Washington, Paper, The World Bank, Washington, D.C. D.C., http://datatopics.worldbank.org/ ______ (2012), “Health Equity and Financial education/. Protection Datasheets”, Latin America and World Bank and ONPES (2014), Haiti. Investing in Sub-Saharan Africa, www.worldbank.org/ people to fight poverty in Haiti: Reflections povertyandhealth . for evidence-based Policy-making (Poverty ______ (2012), “Interim Strategy Note for the Republic Assessment). World Bank Group: of Haiti for FY13-FY14”, Report No. 71885- Washington D.C. HT, The World Bank, Washington, D.C. World Bank, USAID, MSPP (2013), Études sur les ______ (2011), Interim Strategy Note for the Republic ressources humaines et les dépenses de santé of Haiti for CY 2012, Report No. 65112-HT, dans les départements du Nord-Est, Plateau The World Bank, Washington, D.C. Central et Nord-Ouest. ______ (2010), Public Expenditure Tracking and World Economic Forum (2011), The Global Service Delivery Survey – Education and Competitiveness Report 2011-2012, Geneva. Health in Honduras, Latin America and the World Health Organization (2014), Initiative Pour Caribbean Regional Office, The World Bank, un Monde Sans Tabac. Taxation des produits Washington D.C. du tabac, http://www.who.int/tobacco/ ______ (2010), Promoting Nutrition Security in Haiti: economics/taxation/fr/ . An Assessment of Pre- and Post-Earthquake ______ (2014), Initiative Pour un Monde Sans Tabac. Conditions and Recommendations for the Taxation du tabac: un financement novateur Way Forward, The World Bank, Washington, des systèmes de santé, http://www.who.int/ D.C. tobacco/economics/innovative_financing/ ______ (2009), “Burkina Faso: The Budget as fr/ . Centerpiece of PRSP Implementation”, ______ (2014), Haiti: Health Profile, Geneva. 162 BETTER SPENDING, BETTER SERVICES ______ (2014), “Trends In Maternal Mortality: 1990 ______ (2002), World Health Survey, A household to 2013”, Estimates by WHO, UNICEF, questionnaire, World Health Organization, UNFPA, the World Bank and the United Evidence and Information for Policy, the Nations Population Division. World Health Organization, Geneva. ______ (2013), “Enforcing bans on tobacco ______ (2001), “Macroeconomics and Health: advertising, promotion and sponsorship,” Investing in Health for Economic WHO Report on the Global Tobacco Development”, Report of the Commission Epidemic. on Macroeconomics and Health, the World Health Organization, Geneva. ______ (2013), World Health Statistics, the World Health Organization, Geneva. ______, Website, http://www.who.int/nha/country/ hti/en/ ______ (2012), NHA Report 2005-2006, the World Health Organization, Geneva. World Health Organization and the World Bank (2013), Monitoring Progress towards ______ (2011), The Abuja Declaration: Ten Years On. Universal Health Coverage at Country and ______ (2010a), “Monitoring the Building Blocks of Global Levels: A Framework, Joint WHO and Health Systems: A Handbook of Indicators World Bank Group Discussion Paper 1. and Their Measurement Strategies”, The World Vision (2012), Le Système de Protection de World Health Report 2010, Geneva, http:// l’Enfant en Haïti, World Vision, Port au www.who.int/healthinfo/systems/WHO_ Prince, Haiti. MBHSS_2010_full_web.pdf . Xu, K., D. Evans, K. Kawabata, R. Zeramdini, J. ______ (2010b), “Health systems financing: the path Klavus and C. Murray (2003), “Household to universal health coverage,” The World Catastrophic Health Expenditure: A Multi- Health Report 2010, Geneva. Country Analysis.” Lancet 362, pp 111-117. ______ (2010), Rapport sur la Santé dans le Monde, Yves-Francois, P., G. Smucker and J. Tardieu (2009), The World Health Organization, Geneva. “Lost Childhoods in Haiti: Quantifying ______ (2008), Toolkit on Monitoring Health Child Trafficking, Restaveks and Victims Systems Strengthening: Health Systems of Violence”, Final Report, Pan American Financing, http://www.who.int/healthinfo/ Development Foundation. statistics/toolkit_hss/EN_PDF_Toolkit_ Zeng W., M. Cros, K. Dilley, and D. Shepard (2013), HSS_Financing.pdf, the World Health “Impact of performance-based financing on Organization, Geneva. primary health care services in Haiti.” Health ______ (2005), Distribution of health payments and Policy and Planning. catastrophic expenditures methodology, by Ke Zephyr, D., A. Cordova, H. Salgado and M. Seligson Xu. Discussion paper, no 2. (2011), “Haiti in Distress: The Impact of ______ (2004), “The Impact of Health Expenditure the 2010 Earthquake on Citizen Lives on Households and Options for Alternative and Perceptions”, Joint Study by USAID Financing.” Report EM/RC51/4, Regional and Vanderbilt University, http://www. Committee for the Eastern Mediterranean, vanderbilt.edu/lapop/haiti/2010-Haiti-in- the World Health Organization, Cairo, http:// Distress-English.pdf . www.who.int/health_financing/documents/ cov-emrc-healthexpenditureimpact 163 164 BETTER SPENDING, BETTER SERVICES Annexes 165 Annex 1 : PIM Process Flowcharts A. Investment Projects Processing at MPCE DSE 166 B. Investment Projects Processing at MEF DASIP 167 BETTER SERVICES BETTER SPENDING, C. Investment Projects Processing at MEF Treasury 168 D. Investment Projects Processing at Line Ministry: From Request for Funds to Check Issue 169 BETTER SERVICES BETTER SPENDING, E. Investment Projects Processing at Line Ministry: From Check Issue to Payment to Beneficiary 170 BETTER SPENDING, BETTER SERVICES Annex 2 : Correlates with Enrollment School Participation is Unequal-Poverty, Location, Disability, and Other Characteristics Strongly Correlate with Enrollment Dependent Variable: Currently In School Variable Marginal effect Standard error Age 0.034*** (0.011) Gender (1 = male) -0.004 (0.008) Child of household head 0.026* (0.014) Disabled -0.499*** (0.109) Total number of children in household (age 0-18) 0.000 (0.004) Annual hhld consumption per capita (per 1000 gourdes) 0.002*** (0.001) Household head attended but did not complete primary -0.011 (0.013) Household head completed primary 0.038*** (0.013) Household head completed lower secondary or above 0.050*** (0.010) Urban area of residence 0.020* (0.012) Artibonite 0.047*** (0.012) Centre 0.009 (0.018) Grand’Anse 0.025* (0.014) Nippes 0.059*** (0.007) Nord 0.022 (0.014) Nord-Ouest 0.015 (0.021) Ouest 0.031** (0.015) Sud 0.020 (0.016) Sud-Est 0.047*** (0.012) Mean value of dependent variable 0.9065 Observations 4,939 Notes: The regression is estimated for children ages 6-14. Marginal effects evaluated at sample means. Omitted household head education level = no schooling. Omitted department = Nord-Est. Robust standard errors clustered at the household level. *Significant at 10percent. **Significant at 5percent. ***Significant at 1percent. 171 Annex 3 : Correlates with Overage Status Poverty, Location, Gender, and Other Characteristics Strongly Correlate With Overage Status Dependent Variable: At Least 2 Years Over Age For Current Grade Variable Marginal Effect Standard Error Age 0.191 (0.195) Gender (1 = male) 0.079*** (0.028) Child of household head -0.206*** (0.045) Disabled 0.231 (0.167) Total number of children in household (age 0-18) 0.052*** (0.013) Annual hhld consumption per capita (per 1000 gourdes) -0.0030*** (0.001) Household head attended but did not complete primary -0.084** (0.039) Household head completed primary -0.234*** (0.042) Household head completed lower secondary or above -0.285*** (0.045) Urban area of residence -0.192*** (0.036) Artibonite -0.175** (0.074) Centre -0.057 (0.078) Grand’Anse 0.034 (0.077) Nippes -0.105 (0.078) Nord -0.084 (0.074) Nord-Ouest -0.254*** (0.070) Ouest -0.115* (0.066) Sud -0.164** (0.081) Sud-Est -0.040 (0.082) Mean value of dependent variable 0.5165 Observations 2,380 Notes: The regression is estimated for children ages 10-14. Marginal effects evaluated at sample means. Omitted household head education level = no schooling. Omitted department = Nord-Est. Robust standard errors clustered at the household level. *Significant at 10percent. **Significant at 5percent. ***Significant at 1percent. 172 BETTER SPENDING, BETTER SERVICES Annex 4 : EDE PEP Programs: A Glossary EDE PEP has been Haiti’s Government umbrella for centers (implementation starts in 2014-2015). social assistance programs since 2013. Its objective Implementation: MSPP. is to allow beneficiaries to escape extreme poverty • Fight against cholera aims to facilitate the and invest in developing human capital and economic implementation of cholera vaccination health care inclusion. Under EDE PEP, 19 social assistance centers (implementation starts in 2014-2015). programs are in operation through the MAST, FAES, Implementation: MSPP. OFATMA, MSPP, MENFP, PNCS, UCLBP and the MARNDR. • HIMO is a program to develop infrastructures in rural areas (to be determined in 2014-2015). • Bon Solidarité (or Bon Dijan) is a one-time Implementation: MARNDR. cash transfer granted to extreme poor who need immediate intervention. This is an emergency • Kantin Mobil is an in-kind transfer that provides program that targets [assuming it’s still going warm meals to extreme poor living in urban areas. on] households affected by natural disasters.101 Distribution is prioritized for areas that require Implementation: FAES. urgent intervention. Implementation: FAES. • Carte Rose (or Kat Solidarite) is a contributory • Kore Etidyan is a cash transfer targeted to health insurance program that provides access vulnerable families, with a household member to a basic health care package. Implementation: less than 30 years old attending university. OFATMA. Implementation: FAES. • Community health centers. This program aims • Kore Moun Andikape is a monthly cash transfer to implement community pharmacies providing program for individuals between the ages of 18 basic health care (implementation starts in 2014- and 65 living in extreme poverty who have at least 2015). Implementation: MSPP. one disability that prevents them from working, and who receive no subsidy. They must live in • Community restaurant is an in-kind transfer in households, not in disability assistance centers. the form of free meals, provided by 300 restaurants, Implementation: FAES and CAS. to people in disadvantaged neighborhoods in the metropolitan area of Port-au-Prince. ​​ • Kore Peyizan aims to support vulnerable Implementation: FAES (resto PEP) and MAST. households involved in agricultural production by providing a set of agricultural inputs, fertilizers, • Family planning aims to facilitate the etc. Implementation: FAES and MARNDR. establishment of family planning health care • Kore Ti Gran Moun is a cash transfer program that targets households in extreme poverty with 101 Bon Dijans and Panye Solidarite started in a period of an individual over 65 who is not receiving any emergency after Hurricane Sandy. However, the GoH kept them as non-emergency programs for the most vulnerable, hence the change of name from Bon Dijan to Bon Solidarite. 173 pension. Implementation: FAES and MAST • Ranje Kay Kartier/ Banm Lumie-Banm (through the CAS).102 Lavi programs consist of improving urban infrastructures (according to the needs by area). • Literacy [teaching] program (“alphabétisation”) Implementation: UCLBP. targets individuals from 18 to 65 who are illiterate. Implementation: SEA/MENFP. • School Meals [or Canteens] is an in-kind transfer that offers school meals to children during school • Panye Solidarité is an in-kind transfer program days. Implementation: PNCS. that provides a locally produced food kit to households living in extreme poverty, in rural • Ti Kredi offers micro-credits to female-headed areas. This is an emergency program that households engaged in a commercial/productive targets households affected by natural disasters. activity in urban and rural areas. Implementation: Implementation: FAES. FAES. • PSUGO is the program for universal school • Ti Manman Cheri is a conditional cash transfer enrollment. It provides a per-child subsidy to for a mother living in disadvantaged urban areas, public and private schools. Implementation: granted on the condition that her children are MENFP. enrolled in school. Implementation: FAES 102 FAES provides a non-contributory transfer of 400 HTG and MAST provides a pension of 1000 HTG after 20 years of contributions (may be voluntary for informal workers). Harmonization of the program is planned. 174 BETTER SPENDING, BETTER SERVICES Annex 5 : Price Structure for Petroleum Products (2010/2011) 2010-2011 91 95 Diesel Kero PrIx CIF 3.02 3.05 3.01 3.06 Taux annuel moyen 40.66315 40.66315 40.60315 40.66315 Frais financiers 4.50% 4.50% 4.50% 4.50% Total valeur en douane 3.1559 3.187226 3.145475 3.197724 Frais de vérification 5.0.0% 5.o.o.P,6 5.0001i 5.00% Droits de douane 57.80% 57.8oP,6 0.00% 0.00% Frais portuaires 0..32% 0.31% 0.30% 0.29% Droits d’accise 5.40% 5.40% 4.00% 4.00% Accise variable 5.30% 5.25% 3.13% 0.34% Total DTR 73.81% 73.76% 12.43% 9.63% Prix Ex-douane 223.0541 225.198 143.8016 142..S473 Marge compagnie 6.79% 0.082329 0.053ID 0.054132 Marge distributeur 9.02% 10.44% 8.60% 7.61% Transport en province 1.21% 1.20% 1.22% 1.20% MCDT% 17.03% 18.67% 14.97% 13.02% MCDT 21.84831 24.19457 19.15234 16.93213 . DTR + MODT 90.84% 92.43% 21.40% 22.65% Valeur accise Variabfe ($) 0.167228 0.161228 0.098369 0.010821 Pc en $ 6.02 6.13 4.01 3.91 Pc en gdes 244.90 249.39 162.95 159.48 Redevance carburant, gdes 1 1 1 1 Prix final 245.90 250.39 163.95 160.48 Prix à la pompe (Pp) 195 200 162 161 Sources: Direction de l’inspection fiscal, calcul de l’étude Source: IMF 175 176